Cron Context Budget Audit Protocol
Zero-token audit that measures recurring-job prompt and attached-skill static context pressure, detects current context-overflow errors, and separately labels runtime posture (`ACTIVE FAILURE`, `RECOVERED / STRUCTURAL RISK`, `CURRENTLY HEALTHY`, or `UNKNOWN / REVIEW`) so J.A.R.V.I.S. does not confuse a successful recovery run with either an ongoing failure or a structurally fixed payload. It now identifies the dominant payload source, largest-skill share, full attached-skill size breakdown, and a diagnostic classification projection without the largest skill; the projection is explicitly evidence-only and never authorizes removing a required skill. Gives a repair-before-rerun cue without editing or changing job state; integrated into the proactive commander cycle and evidence freshness index.
Status: active
Location: /opt/data/agent-bridge/CRON_CONTEXT_BUDGET_AUDIT_PROTOCOL.md`; helper at `/opt/data/profiles/dev/scripts/cron-context-budget-audit.py`; latest artifacts at `/opt/data/agent-bridge/recurring-jobs/context-budget-audit-latest.md` and `.json
Cron Context Compaction Preview
Zero-token, non-mutating builder that compares the live D.E.V. proactive-loop prompt and attached skills with a compact canonical-protocol proposal, measures projected context reduction, verifies referenced protocol paths and retained skill availability, preserves schedule/delivery/toolset/model/provider evidence, and gives J.A.R.V.I.S. a decision-ready preview without editing or rerunning the cron job. Current/proposed prompt+skills SHA-256 fingerprints enforce a compare-and-swap stale-preview guard so a future approved edit must stop if live scoped state drifted after review. An in-memory apply rehearsal must also prove that only `prompt` and `skills` change, the source job remains unchanged, no unexpected fields drift, and the candidate fingerprint matches the reviewed proposal before the preview can return `READY FOR APPROVAL`. Integrated into the proactive commander cycle and commander evidence freshness index.
Status: active
Location: /opt/data/profiles/dev/scripts/cron-context-compaction-preview.py`; latest artifacts at `/opt/data/agent-bridge/recurring-jobs/context-compaction-preview-latest.md`, `.json`, and `.txt
No Reinventing the Wheel — Folder + Source-of-Truth Protocol
Requires agents to check existing proper folders, docs, sheets, databases, protocols, and source-of-truth spaces before creating new ones. Agents must reuse/update/merge where possible, create new only when justified, use canonical agent folder names such as `A.C.E.`, and include a no-reinvent proof line in bridge reports. Installed after correcting the duplicate `ACE` folder into the proper `A.C.E.` structure.
Status: active
Location: /opt/data/agent-bridge/team/NO_REINVENT_THE_WHEEL_FOLDER_AND_SOURCE_OF_TRUTH_PROTOCOL.md
Driving Mode Protocol
Hands-free, low-distraction safety mode for when Stufio is driving or cannot safely read/type. Requires concise audio-friendly responses, defers visual/typing-heavy steps until parked, prevents noisy agent chatter, preserves Slack-only/Telegram-paused operations, and keeps all normal approval gates intact.
Status: active
Location: /opt/data/agent-bridge/DRIVING_MODE_PROTOCOL.md
Iris Confidence-Tagged Action Packaging Protocol
Requires each Iris-packaged action to carry a `high` / `medium` / `low` confidence tag, evidence-based reason, concrete follow-up gate, approval sensitivity, and stop path before J.A.R.V.I.S. routes it as active work.
Status: active
Location: /opt/data/agent-bridge/IRIS_CONFIDENCE_TAGGED_ACTION_PACKAGING_PROTOCOL.md`; template at `/opt/data/agent-bridge/iris-jarvis/templates/confidence-tagged-action-package.md`; first rehearsal packet at `/opt/data/agent-bridge/dev-jarvis/20260719T060904Z-iris-confidence-tagged-action-packaging.md
D.E.V. Proactive Improvement Protocol
Defines how D.E.V. proactively improves J.A.R.V.I.S., current agents, future agents, documentation, workflows, and internal systems while respecting approval boundaries. Requires night-and-day material advancement rather than monitoring-only work, with J.A.R.V.I.S. leadership/delegation intelligence treated as a primary upgrade lane every cycle, and now explicitly requires `Status:` plus `Improvements Made:` fields in automatic reports.
Status: active
Location: /opt/data/agent-bridge/DEV_proactive_improvement_protocol.md
Proactive WATCH Pattern Ledger Protocol
Zero-token local ledger that scans recent D.E.V. → J.A.R.V.I.S. proactive WATCH recovery packets, identifies repeated first-approval asks or recurring hold patterns, and gives J.A.R.V.I.S. a compact dedupe/trend cue before live remediation is considered. The commander-cycle builder now refreshes the decision matrix, decision pack/aging audit, and first-ask trend before this ledger runs, preventing the ledger from anchoring a superseded approval priority. It separately anchors the current first approval ask from `/opt/data/agent-bridge/approval-queue/first-approval-ask-latest.txt`; approval lines, safety boundaries, and stop paths must render without ellipsis truncation in commander-facing artifacts.
Status: active
Location: /opt/data/agent-bridge/PROACTIVE_WATCH_PATTERN_LEDGER_PROTOCOL.md`; helper at `/opt/data/profiles/dev/scripts/proactive-watch-pattern-ledger.py`; latest artifacts at `/opt/data/agent-bridge/proactive-loop/watch-pattern-ledger-latest.md`, `.json`, and `.txt
Approval Item Relevance Ordering Protocol
Prevents old approval items from becoming J.A.R.V.I.S.'s first ask when verified current evidence shows their preparatory work is complete, their target remains safely held, or a separate pending item now owns the remaining live decision. Adds explicit `CURRENT`, `STALE-REVIEW`, and `SUPERSEDED-REVIEW` relevance markers; current decisions rank before review-marked items, while review markers never approve, reject, resolve, delete, or execute an item. Profile-specific repair items must now be reconciled against the latest deterministic finding set so a repaired/absent target cannot outrank current profile findings.
Status: active
Location: /opt/data/agent-bridge/APPROVAL_ITEM_RELEVANCE_ORDERING_PROTOCOL.md`; helper at `/opt/data/profiles/dev/scripts/proactive-improvement-decision-matrix.py`; latest artifact at `/opt/data/agent-bridge/proactive-loop/proactive-improvement-decision-matrix-latest.md
Approval First-Ask Operator Card Protocol
Keeps a recurring top approval ask synchronized with the current first-ask title, pending count, exact approval line, post-decision options, and commander evidence freshness. The helper requires current-title alignment and freshness PASS, then fails closed without overwriting the scoped card if commander priority changes.
Status: active
Location: /opt/data/agent-bridge/APPROVAL_FIRST_ASK_OPERATOR_CARD_PROTOCOL.md`; helper at `/opt/data/profiles/dev/scripts/approval-first-ask-operator-card-refresh.py`; current artifacts at `/opt/data/agent-bridge/approval-queue/nimbus-first-ask-operator-card-latest.md`, `.json`, and `.txt
D.E.V. Checkpoint Summary Protocol
Defines the required checkpoint update structure and governance for concise continuity handoffs, including Active Task, Goal, Constraints & Preferences, Completed Actions, Active State, In Progress, Blocked, Key Decisions, Resolved Questions, Pending User Asks, Relevant Files, Remaining Work, and Critical Context. Emphasizes preserving relevant history, removing only obsolete content, and masking secrets.
Status: active
Location: /opt/data/agent-bridge/DEV_CHECKPOINT_SUMMARY_PROTOCOL.md
Night-and-Day System Advancement Protocol
Converts Stufio's correction into an active rule that proactive D.E.V. work must produce visible, compounding system upgrades instead of only monitoring or small status checks. Requires material advancements, daily leap domains, interruption handling, report fields, approval boundaries, a standing J.A.R.V.I.S. Leadership Check, a standing Central Command Check, and mandatory owning-domain-agent consultation before/during technical work (A.C.E. first for production/events/Equipment Hub).
Status: active
Location: /opt/data/agent-bridge/NIGHT_AND_DAY_SYSTEM_ADVANCEMENT_PROTOCOL.md
Night-and-Day System Advancement Backlog
Prioritized backlog of safe and approval-needed system leaps, including Command Center evolution, agent upgrade cards, mission-to-action conversion, A.C.E. production OS, N.E.X.U.S. opportunity engine, and larger approval-needed integrations. Now includes clearer status/progress, verification/evidence, and approval/safety metadata so J.A.R.V.I.S. can distinguish safe internal leaps from approval-gated concepts before routing work.
Status: active
Location: /opt/data/agent-bridge/NIGHT_AND_DAY_SYSTEM_ADVANCEMENT_BACKLOG.md
Future Agent Readiness Bootstrap Protocol
Minimal read-only bootstrap pattern for any future or dormant agent before routing, recurring jobs, tool access growth, or authority expansion. Defines required roster/protocol inheritance checks, readiness scorecard commands, explicit HOLD/candidate/limited/expansion classes, and evidence packet fields.
Status: active
Location: /opt/data/agent-bridge/FUTURE_AGENT_READINESS_BOOTSTRAP_PROTOCOL.md
Future Agent Quickstart Brief Protocol
Standard one-screen read-only protocol for future lanes transitioning from concept/dormant to active consideration. Requires a quickstart brief with decision cue, required gates, approval sensitivity, next safe action, and stop path before authority expansion; now also includes confidence tier and J.A.R.V.I.S. leadership/coordination impact fields.
Status: active
Location: /opt/data/agent-bridge/FUTURE_AGENT_QUICKSTART_BRIEF_PROTOCOL.md
System Advancement Backlog Health Audit Script
Zero-token local audit that checks the two proactive advancement backlogs for commander-usable metadata: status/progress, approval/safety boundary, verification/evidence, and a concrete smallest-safe-step or upgrade. It gives J.A.R.V.I.S. a PASS/WATCH/ATTENTION cue before turning backlog ideas into delegated work or approval packets.
Status: active
Location: /opt/data/profiles/dev/scripts/system-advancement-backlog-health-audit.py`; latest artifacts at `/opt/data/agent-bridge/system-advancement/backlog-health-audit-latest.json` and `.md
Night-and-Day Wow Delta Digest Builder
Zero-token local builder that summarizes recent entries in the agent skill/procedure improvement log into a commander-readable Night-and-Day / Wow Delta digest, including improvement counts, active dates, impact tags, verification excerpts, changed paths, safety posture, and next safe visibility upgrade.
Status: active
Location: /opt/data/profiles/dev/scripts/wow-delta-digest-builder.py`; latest artifacts at `/opt/data/agent-bridge/system-advancement/wow-delta-digest-latest.json` and `.md
J.A.R.V.I.S. Leadership Upgrade Protocol
Standing protocol requiring D.E.V.'s recurring improvement work to upgrade or evaluate J.A.R.V.I.S.'s ability to lead the team, delegate, give proper instructions, analyze agent activity, improve future instructions, respond to team decision requests, run a recurring leadership pulse, proactively execute safely, and communicate useful observations to Stufio. It now standardizes leader continuity with measurable lanes and links the Leadership Pulse protocol for measurable command feedback across cycles.
Status: active
Location: /opt/data/agent-bridge/JARVIS_LEADERSHIP_UPGRADE_PROTOCOL.md
J.A.R.V.I.S. Leadership Pulse Protocol
Lightweight recurring measurement protocol for commander leadership quality. Defines 5 leadership lanes (command clarity, delegation quality, responsiveness, safe execution discipline, future-agent inclusion), scoring rules, pulse cues, and anti-regression guards so J.A.R.V.I.S. can prove measurable leadership progress over cycles. Adds readiness-coupled pulse fields and a standard one-line capture for bold leadership experiment ideas before they become unsafe live changes.
Status: active
Location: /opt/data/agent-bridge/JARVIS_LEADERSHIP_PULSE_PROTOCOL.md
J.A.R.V.I.S. Leadership + Delegation Packet Protocol
One-screen commander pre-brief that combines the latest leadership pulse, delegation debt, readiness index status, approval-gated count, active mission-action count, current first move, first safe action, and exact first approval line so J.A.R.V.I.S. can decide whether to hold, ask, or advance without hunting across multiple ledgers.
Status: active
Location: /opt/data/agent-bridge/JARVIS_LEADERSHIP_DELEGATION_PACKET_PROTOCOL.md`; helper at `/opt/data/profiles/dev/scripts/jarvis-leadership-delegation-packet.py`; latest artifacts at `/opt/data/agent-bridge/commander/leadership-delegation-packet-latest.md`, `.json`, and `.txt
J.A.R.V.I.S. Delegation Debt Report Protocol
Internal protocol for adding a one-screen delegation debt block to proactive reports when repeated assignment misses appear. Standardizes a short summary of leadership debt, a next mitigation step, and a validation cue so J.A.R.V.I.S. can improve instruction quality without changing live routing or permissions.
Status: active
Location: /opt/data/agent-bridge/JARVIS_DELEGATION_DEBT_REPORT_PROTOCOL.md
J.A.R.V.I.S. Leadership Pulse Snapshot Helper
Internal helper to generate a quick deterministic five-lane Leadership Pulse draft snapshot from bridge artifacts (`approval_queue.md`, `mission_action_queue.md`, leadership and readiness docs), intended as a one-screen, safe pre-work input to the canonical pulse note.
Status: active
Location: /opt/data/agent-bridge/scripts/leadership_pulse_snapshot.py
J.A.R.V.I.S. Instruction Quality Gate Protocol
10-point pre-send gate for meaningful specialist assignments and recurring/future-agent work, checking outcome, active owner, role fit, need-to-know context, scope, approval gates, inputs/dependencies, verification, cadence, and stop path before J.A.R.V.I.S. delegates. Now links the Delegation Rewrite Playbook so sub-8/10 instructions are rewritten before dispatch or re-dispatch.
Status: active
Location: /opt/data/agent-bridge/JARVIS_INSTRUCTION_QUALITY_GATE_PROTOCOL.md
J.A.R.V.I.S. Delegation Rewrite Playbook
Concrete weak-instruction to strong-instruction rewrite patterns for vague outcomes, inactive owners, missing approval gates, missing verification, missing stop paths, overshared context, and monitoring-only improvement loops. Used when an assignment scores below 8/10 or specialist output reveals commander-side instruction weakness.
Status: active
Location: /opt/data/agent-bridge/JARVIS_DELEGATION_REWRITE_PLAYBOOK.md
Agent Delegation Intake Packet Protocol
Lightweight pre-flight packet J.A.R.V.I.S. uses before routing work to a specialist agent, recurring job, or future-agent lane; requires mission, active owner check, scope, approval gates, inputs/dependencies, deliverable, verification, reporting cadence, and stop path so weak assignments do not become duplicated work, silence, or unsafe expansion.
Status: active
Location: /opt/data/agent-bridge/AGENT_DELEGATION_INTAKE_PACKET_PROTOCOL.md
Mission-to-Action Conversion Protocol
Converts Stufio directives, Slack/voice requests, bridge notes, recurring-job findings, and proactive-loop discoveries into concrete owner-ready action records with source, owner, scope, approval gates, verification, evidence path, and stop path so mission work does not remain loose chat context. Now names the bridge-first mission action queue as the default persistent tracking surface when the Command Center app action store is not being used.
Status: active
Location: /opt/data/agent-bridge/MISSION_TO_ACTION_CONVERSION_PROTOCOL.md
Mission Action Queue
Bridge-first action ledger for persistent owner-ready mission records when a Command Center app action record is not yet created; seeded with current provider-failure and Slack-delivery triage actions from the D.E.V. proactive health precheck. Now linked to the Mission Action Queue Health Audit Script for zero-token checks before J.A.R.V.I.S. relies on assignment or completion state.
Status: active
Location: /opt/data/agent-bridge/mission_action_queue.md
Mission Action Queue Health Audit Script
Zero-token read-only scanner that checks `/opt/data/agent-bridge/mission_action_queue.md` for active action required fields, approved status labels, commander cue labels, evidence paths, approval-gate handling, and completed-action evidence/verification/safety notes so J.A.R.V.I.S. can trust the bridge-first mission queue before assigning, closing, or escalating follow-up work.
Status: active
Location: /opt/data/profiles/dev/scripts/mission-action-queue-health-audit.py`; latest artifacts at `/opt/data/agent-bridge/mission-action/queue-health-audit-latest.json` and `.md
Commander Brief Pack Protocol
Compact decision-ready brief format that compresses status, changed state, verification, blockers, decision needed, recommendation, and next safe action into one screen so J.A.R.V.I.S. can lead and decide faster without noise.
Status: active
Location: /opt/data/agent-bridge/COMMANDER_BRIEF_PACK_PROTOCOL.md
D.E.V. Proactive Loop Commander Cycle Builder
Zero-token local builder that refreshes the safe commander-cycle evidence chain and writes one compact packet showing fresh surfaces, approval-gated items, and hold boundaries before live action. Each run stores timestamped history snapshots, refreshes approval, Central Command, recurring-job, readiness, leadership, evidence-hygiene, and next-moves surfaces, then regenerates the proactive composite packet after the current cycle and next-moves artifacts exist. Final freshness scoring runs after the composite write so J.A.R.V.I.S. cannot receive a stale one-screen first-read packet beside an otherwise fresh commander cycle.
Status: active
Location: /opt/data/profiles/dev/scripts/proactive-loop-commander-cycle-builder.py`; latest artifacts at `/opt/data/agent-bridge/proactive-loop/commander-cycle-latest.json` and `.md
D.E.V. Proactive Loop Commander Cycle Trend Audit
Zero-token local audit that reads timestamped commander-cycle history snapshots, counts recurring PASS/WATCH/ATTENTION patterns, tracks approval-decision counts, and identifies repeated non-OK commander surfaces so J.A.R.V.I.S. can distinguish one-off warnings from trend evidence before live routing, restart, job-state, provider, or approval decisions.
Status: active
Location: /opt/data/profiles/dev/scripts/proactive-loop-commander-cycle-trend-audit.py`; latest artifacts at `/opt/data/agent-bridge/proactive-loop/commander-cycle-trend-audit-latest.json` and `.md
J.A.R.V.I.S. Commander Brief Pack Builder
Zero-token local builder that condenses decision surface, evidence freshness, wow delta, proactive precheck, and approval decision pack into the Commander Brief Pack labels so J.A.R.V.I.S. has one first-read leadership brief before routing, approval, restart, job-state, or future-agent decisions. Also writes a clean Slack-visible `.txt` artifact with no raw markdown emphasis, code ticks, or pipe-table characters.
Status: active
Location: /opt/data/profiles/dev/scripts/commander-brief-pack-builder.py`; latest artifacts at `/opt/data/agent-bridge/commander/brief-pack-latest.json`, `.md`, and `.txt
J.A.R.V.I.S. Commander Next Moves Builder
Zero-token local builder that condenses the commander cycle, trend audit, commander brief, approval decision pack, proactive improvement decision matrix, decision surface, and evidence freshness into an ordered next-moves brief. It separates approval-gated live actions from safe internal refreshes, aligns the first approval ask with the current decision-matrix priority order, preserves timestamped snapshots for trend comparison, keeps approval-adjacent safety boundaries untruncated so commander decision packets do not hide stop lines, normalizes repeated safety field labels in next-move rows without altering exact approval lines, and emits a clean Slack-readable `.txt` artifact with no raw markdown emphasis, code ticks, or pipe-table characters.
Status: active
Location: /opt/data/profiles/dev/scripts/commander-next-moves-builder.py`; latest artifacts at `/opt/data/agent-bridge/commander/next-moves-latest.json`, `.md`, and `.txt`; history snapshots under `/opt/data/agent-bridge/commander/next-moves/`.
J.A.R.V.I.S. Commander Next-Moves Trend Audit
Zero-token local audit that reads timestamped commander next-moves snapshots, counts repeated approval-gated move titles, tracks approval-gated and safe-internal series, surfaces the latest first approval-gated move with its exact approval line/safety boundary, and gives J.A.R.V.I.S. a commander cue when the same Stufio/J.A.R.V.I.S. decisions recur across proactive-loop runs.
Status: active
Location: /opt/data/profiles/dev/scripts/commander-next-moves-trend-audit.py`; latest artifacts at `/opt/data/agent-bridge/commander/next-moves-trend-audit-latest.json` and `.md`; source history under `/opt/data/agent-bridge/commander/next-moves/`.
Slack-Readable Artifact Format Audit Protocol
Zero-token/local audit that checks known Slack-visible `.txt` commander, proactive, Central Command profile-normalization, N.E.X.U.S. review, and approval briefs plus generated decision-card and operator-card text surfaces for Stufio's formatting rule: first line identifies D.E.V., no raw backticks, asterisks, or pipe-table characters appear, and approval/safety labels are not duplicated before J.A.R.V.I.S./Stufio treat the artifact as Slack-ready. New first-read `.txt` packet families must be added to this deterministic audit in the same implementation pass.
Status: active
Location: /opt/data/agent-bridge/SLACK_READABLE_ARTIFACT_FORMAT_AUDIT_PROTOCOL.md`; helper at `/opt/data/profiles/dev/scripts/slack-readable-artifact-format-audit.py`; latest artifacts at `/opt/data/agent-bridge/slack-readable-artifacts/format-audit-latest.json` and `.md
Commander Quality Gate Checklist
Fast commander gate for deciding whether specialist output is ready for Stufio, should be sent back, is blocked, or needs approval; checks request fit, evidence, safety, presentation quality, Stufio preference alignment, and next-action clarity.
Status: active
Location: /opt/data/agent-bridge/COMMANDER_QUALITY_GATE_CHECKLIST.md
J.A.R.V.I.S. Decision Cue Protocol
Standardizes the one-line next-action cue for commander surfaces, mapping nominal, ready, overloaded, blocked, escalation-needed, approval-needed, and uncertain states into concise operational verbs so J.A.R.V.I.S. can choose the safest next move at a glance.
Status: active
Location: /opt/data/agent-bridge/JARVIS_DECISION_CUE_PROTOCOL.md
J.A.R.V.I.S. Commander Decision Surface Builder
Zero-token local builder that condenses approval decisions, active mission actions, recurring-job WATCH/HOLD items, and current audit cues into one commander-readable decision surface so J.A.R.V.I.S. can see what needs approval, triage, or safe internal continuation without rereading multiple ledgers.
Status: active
Location: /opt/data/profiles/dev/scripts/commander-decision-surface-builder.py`; latest artifacts at `/opt/data/agent-bridge/commander/decision-surface-latest.json` and `.md
J.A.R.V.I.S. Commander Evidence Freshness Index
Zero-token local index that checks whether the commander evidence surfaces J.A.R.V.I.S. relies on are fresh, stale, or missing before live routing, resume/rerun, approval, restart, Central Command, future-agent activation, or agent capability-card decisions. Includes Agent Upgrade Card review-audit freshness and the proactive composite packet so neither an overdue deck nor a stale one-screen first-read packet can remain outside the main commander gate; clamps same-second negative age drift to `0.0`.
Status: active
Location: /opt/data/profiles/dev/scripts/commander-evidence-freshness-index.py`; latest artifacts at `/opt/data/agent-bridge/commander/evidence-freshness-index-latest.json` and `.md
J.A.R.V.I.S. Team Response Obligation Protocol
Standing rule that J.A.R.V.I.S. must answer team requests for decisions, routing calls, approval interpretation, commander instructions, chair actions, or ownership choices instead of leaving delegated work stuck in silence; includes decision/delegation/blocker/escalation/hold response modes.
Status: active
Location: /opt/data/agent-bridge/JARVIS_TEAM_RESPONSE_OBLIGATION_PROTOCOL.md
Central Command Center Night-and-Day Improvement Protocol
Standing rule that every 30-minute D.E.V. improvement loop must improve, verify, or approval-package the shared Slack command room so group meetings run smoothly with J.A.R.V.I.S. chairing, one-speaker floor control, explicit roundtables, no duplicate/echo replies, scorecards, and actual evidence before over-reporting. Now includes conservative WATCH chairing mode so WATCH/ATTENTION evidence becomes named-floor control and exact approval-first behavior instead of live execution drift.
Status: active
Location: /opt/data/agent-bridge/CENTRAL_COMMAND_CENTER_NIGHT_AND_DAY_PROTOCOL.md
Central Command Center Improvement Backlog
Active backlog of Central Command improvements including specialist free-response normalization, J.A.R.V.I.S. chair packet, room smoothness scorecard, human-triggered room drill, and future Command Center app integration surface. Now includes explicit approval/safety and verification metadata for each Central Command backlog item so room upgrades are not mistaken for authorization to edit routing or restart gateways.
Status: active
Location: /opt/data/agent-bridge/CENTRAL_COMMAND_CENTER_IMPROVEMENT_BACKLOG.md
Central Command Drill Result Capture Template
Secret-free template for recording live Central Command drill results after Stufio/J.A.R.V.I.S. triggers the room, including safety pre-check, named-agent floor test, silence/no-open-floor test, J.A.R.V.I.S. roundtable chair test, parking-lot capture test, score, evidence path, next fix, and approval-needed line.
Status: active
Location: /opt/data/agent-bridge/CENTRAL_COMMAND_DRILL_RESULT_CAPTURE_TEMPLATE.md
Central Command Live Drill Prompt Pack
Human-triggered drill script pack that turns the pending Central Command approval item into exact Slack prompts for named-agent floor control, silence/no-open-floor behavior, J.A.R.V.I.S. roundtable chairing, and parking-lot capture, with evidence paths, safety boundaries, and a one-screen operator card for moving from approval to captured evidence without live routing or gateway changes.
Status: active
Location: /opt/data/agent-bridge/CENTRAL_COMMAND_LIVE_DRILL_PROMPT_PACK.md
Central Command Drill Evidence Index Script
Zero-token read-only scanner that indexes local Central Command drill support artifacts and distinguishes supporting runbooks/tests from actual live drill result evidence, so J.A.R.V.I.S. does not claim room verification or approve routing/job changes before human-triggered evidence exists.
Status: active
Location: /opt/data/profiles/dev/scripts/central-command-drill-evidence-index.py`; latest artifacts at `/opt/data/agent-bridge/central-command/drill-evidence-index-latest.json` and `.md
Central Command Room Smoothness Audit Script
Zero-token read-only scanner that checks active-agent Slack config posture for Central Command room smoothness: specialist free-response drift, require-mention posture, channel prompts, name triggers, ignore triggers, and commander cue before any live routing/profile/gateway change.
Status: active
Location: /opt/data/profiles/dev/scripts/central-command-room-smoothness-audit.py`; latest artifacts at `/opt/data/agent-bridge/central-command/room-smoothness-audit-latest.json` and `.md
Central Command Slack Send Failure Response Protocol
One-screen read-only recovery protocol for recurring-loop states where Slack send failures (including interpreter-shutdown and future send transport errors) must be triaged before any resume/unpause or routing change. It standardizes the watch-only evidence stack, blocker classing, and recovery packet outputs.
Status: active
Location: /opt/data/agent-bridge/CENTRAL_COMMAND_SLACK_SEND_FAILURE_RESPONSE_PROTOCOL.md
Central Command Profile Normalization Decision Packet Protocol
Converts profile-specific Central Command room-smoothness WATCH/ATTENTION findings into profile-by-profile decision lanes: J.A.R.V.I.S. chairing cue, documentation-safe packet, or approval-gated repair with exact verification and stop paths before live routing/profile/gateway changes. The helper now regenerates the packet deterministically from the latest room-smoothness audit and writes a clean Slack-readable text sibling for J.A.R.V.I.S. first-read use.
Status: active
Location: /opt/data/agent-bridge/CENTRAL_COMMAND_PROFILE_NORMALIZATION_DECISION_PACKET_PROTOCOL.md`; helper at `/opt/data/profiles/dev/scripts/central-command-profile-normalization-decision-packet.py`; template at `/opt/data/agent-bridge/central-command/templates/profile-normalization-decision-packet.md`; latest artifacts at `/opt/data/agent-bridge/central-command/profile-normalization-decision-packet-latest.md`, `.json`, and `.txt
Central Command Chair Packet Builder Script
Zero-token local builder that combines Central Command smoothness audit evidence, live-drill evidence status, approval decision pack items, and commander decision surface cues into a one-screen J.A.R.V.I.S. chair packet with floor, queue, decision-needed, parking-lot, next-action, and safety-boundary fields.
Status: active
Location: /opt/data/profiles/dev/scripts/central-command-chair-packet-builder.py`; latest artifacts at `/opt/data/agent-bridge/central-command/chair-packet-latest.json` and `.md
Delegated Decision Reporting Protocol
Defines when D.E.V. must report an owning agent's response back to Stufio after Stufio delegates a decision, versus normal internal coordination where D.E.V. should let J.A.R.V.I.S./the owning agent decide and then safely implement without unnecessary interruption.
Status: active
Location: /opt/data/agent-bridge/DELEGATED_DECISION_REPORTING_PROTOCOL.md
Agent Upgrade Cards
Concrete per-agent upgrade cards defining next capabilities, safe internal upgrades, first prototypes, and approval gates so proactive improvement spans all active and future agent lanes. The deck received a documentation-only evidence review on 2026-07-25: all 10 cards retained their existing approval gates and stop paths, and review dates were moved to a short accountable window. Current review evidence is at `/opt/data/agent-bridge/agent-upgrade-cards/review-audit-latest.md`.
Status: active
Location: /opt/data/agent-bridge/AGENT_UPGRADE_CARDS.md
M.A.C. GUI Verification Checklist Protocol
Internal verification protocol and template that require a confidence rating, acceptance test, evidence path, rollback/failure note, stop path, and approval-needed flag before local GUI or physical-task support work is reported as complete.
Status: active
Location: /opt/data/agent-bridge/MAC_GUI_VERIFICATION_CHECKLIST_PROTOCOL.md`; template at `/opt/data/agent-bridge/mac-jarvis/templates/gui-verification-checklist.md
Equipment Hub Starter Kit
Internal mock-safe equipment operating template defining core item fields, readiness checks, audit views, safe mock import columns, mock CSV validation preview, local readiness scorecard, and approval gates before live inventory, rental, vendor, purchase, or production workflows.
Status: active
Location: /opt/data/agent-bridge/equipment-hub/templates/equipment-hub-starter-kit.md
Agent Improvement Request Inbox Protocol
Shared place for J.A.R.V.I.S., A.C.E., D.E.V., and future agents to record capability requests such as “D.E.V., I would be better if I could ___.” D.E.V. evaluates usefulness, risk, approval requirements, safe implementation path, and now links every review to the Agent Improvement Request Triage Protocol so raw requests become clear commander routes.
Status: active
Location: /opt/data/agent-bridge/agent_improvement_request_inbox.md
Agent Improvement Request Triage Protocol
Converts agent “I would be better if…” requests into one of five commander-ready paths: safe internal improvement, mission action, readiness input, approval-needed item, or hold/clarity request, with required fields, recurrence handling, a reusable one-screen evidence block, safety gates, and verification.
Status: active
Location: /opt/data/agent-bridge/AGENT_IMPROVEMENT_REQUEST_TRIAGE_PROTOCOL.md
Agent Improvement Inbox Audit Script
Zero-token local audit that reads `/opt/data/agent-bridge/agent_improvement_request_inbox.md`, confirms whether pending agent improvement requests exist, flags untriaged requests, and writes a commander-readable cue so J.A.R.V.I.S. knows when capability requests need routing instead of letting raw notes go stale.
Status: active
Location: /opt/data/profiles/dev/scripts/agent-improvement-inbox-audit.py`; latest artifacts at `/opt/data/agent-bridge/agent-improvement/inbox-audit-latest.json` and `.md
Protocol Registry Maintenance Protocol
Requires D.E.V. to maintain this central list of all protocols and their descriptions so Stufio can request a protocol report/sheet at any time.
Status: active
Location: /opt/data/agent-bridge/protocol_registry.md
Protocol Registry Health Audit Script
Zero-token local audit that scans `/opt/data/agent-bridge/protocol_registry.md`, confirms registry entries keep required fields, detects duplicate headings, flags malformed list-prefix drift, ignores deliberate state-marker absence such as an inactive emergency brake file, and writes commander-readable health artifacts so J.A.R.V.I.S. can trust the protocol registry as a routing/reporting source.
Status: active
Location: /opt/data/profiles/dev/scripts/protocol-registry-health-audit.py`; latest artifacts at `/opt/data/agent-bridge/protocol-registry/health-audit-latest.json` and `.md
Team Agent Safety Protocol
Defines command hierarchy, need-to-know privacy, specialist-agent safety responsibilities, high-risk actions requiring explicit Stufio approval, escalation language, and safety review cadence.
Status: active
Location: /opt/data/agent-bridge/TEAM_AGENT_SAFETY_PROTOCOL.md
Team Agent Memory & Interconnection Protocol
Defines how agents inherit shared coordination rules, bridge-folder patterns, update acknowledgements, and disciplined memory use while preserving need-to-know boundaries.
Status: active
Location: /opt/data/agent-bridge/TEAM_AGENT_MEMORY_PROTOCOL.md
Initiate Recalibrate Protocol
Standing recovery command triggered by `Initiate recalibrate <agent>` that forces the named agent to restore identity/persona, safety rules, memory/context, bridge notes, scheduled duties, active obligations, blockers, and approval gates from durable sources; includes relevance/ownership filters, overlap control, source-quality checks, context classification, and pass/fail verification before reporting back with evidence.
Status: active
Location: /opt/data/agent-bridge/INITIATE_RECALIBRATE_PROTOCOL.md
Agent Capability Matrix & Safe Expansion Map
Internal capability map for active agents, future-agent candidate lanes, expansion gates, and safe experiment backlog entries so J.A.R.V.I.S. can route work and D.E.V. can plan growth without granting unsafe authority by accident. Adds a Delegation & execution dispatch specialist lane, an explicit Nimbus activation gate experiment for unresolved candidate lanes, and a Readiness Delta Steward experiment to turn readiness drift into a cross-agent operator cue.
Status: active
Location: /opt/data/agent-bridge/AGENT_CAPABILITY_MATRIX.md
Agent Upgrade Card Protocol
Defines the upgrade-card operating language for each agent lane, including next capability leap, approval sensitivity, safety boundaries, evidence anchors, and next safe action. Includes a companion deck file (`/opt/data/agent-bridge/AGENT_UPGRADE_CARDS.md`) plus a zero-token review-date audit now integrated into the proactive commander cycle, evidence freshness index, and composite first-read packet so overdue cards cannot silently drive assignments or authority decisions.
Status: active
Location: /opt/data/agent-bridge/AGENT_UPGRADE_CARD_PROTOCOL.md`; audit helper `/opt/data/profiles/dev/scripts/agent-upgrade-card-review-audit.py`; latest artifacts `/opt/data/agent-bridge/agent-upgrade-cards/review-audit-latest.md` and `.json
Proactive Watch-to-Action Steward Protocol
Defines a no-blind-retry steward packet workflow that turns proactive precheck/decision-matrix/watch signals into a reversible, one-page operator-safe action packet before any job-state, routing, provider, or gateway changes are made.
Status: active
Location: /opt/data/agent-bridge/PROACTIVE_WATCH_TO_ACTION_STEWARD_PROTOCOL.md
Active Agent Roster & Activation Protocol
Canonical internal roster for active agents, inactive/future-agent lanes, routing rules, and the checklist required before a new agent is treated as active.
Status: active
Location: /opt/data/agent-bridge/ACTIVE_AGENT_ROSTER.md
Create Agent Protocol
Triggered when Stufio asks to create a new agent; defines the initial responsibility-capture question, dotted acronym/name rules, personality creation, Hermes profile/SOUL/bridge setup, safety and memory inheritance, profile identity test, and messaging-credential gates.
Status: active
Location: /opt/data/agent-bridge/CREATE_AGENT_PROTOCOL.md
Future Agent Onboarding Checklist
Safe minimum onboarding checklist for future agents covering identity, hierarchy, safety, memory/privacy, bridge setup, protocol registry linkage, least-privilege tools/access, verification, post-action reporting, approval queue/ledger handling, readiness scorecard and sandbox-drill inheritance, continuity handoff packets, activation acceptance packets, operating-tier labels, zero-token activation-gate audit, packet template/scaffold linkage, and one-screen commander handoff to J.A.R.V.I.S.
Status: active
Location: /opt/data/agent-bridge/FUTURE_AGENT_ONBOARDING_CHECKLIST.md
Future Agent Activation Acceptance Packet Template and Scaffold
Standard required-field template and safe-by-default scaffold for future-agent activation acceptance packets, covering operating tier, domain owner, bridge/profile/messaging status, Slack-only/Telegram-paused check, free-response posture, readiness/drill/resume evidence, approval gates, stop path, and J.A.R.V.I.S. commander cue before any ACTIVE roster promotion.
Status: active
Location: /opt/data/agent-bridge/FUTURE_AGENT_ACTIVATION_ACCEPTANCE_PACKET_TEMPLATE.md`; scaffold script at `/opt/data/profiles/dev/scripts/future-agent-activation-packet-scaffold.py
Agent Sandbox Drill Protocol
No-live-execution drill pattern used before expanded responsibility, credentials/tools/access, recurring autonomy, public/customer-facing workflows, production authority, or sensitive integrations; includes identity/routing, safety/approval, tool dry-run, reporting, and coordination drill types plus evidence requirements.
Status: active
Location: /opt/data/agent-bridge/AGENT_SANDBOX_DRILL_PROTOCOL.md
Approval Queue Protocol
Defines the `⚠️ APPROVAL NEEDED` marker and approval queue workflow so Stufio can quickly approve, reject, or revise ideas/actions that need attention while D.E.V. continues safe autonomous improvements.
Status: active
Location: /opt/data/agent-bridge/approval_queue_protocol.md
D.E.V. Approval Queue
Living list of approval-needed items using the `⚠️ APPROVAL NEEDED` marker, with proposed action, value, risk notes, options, and D.E.V. recommendation.
Status: active
Location: /opt/data/agent-bridge/approval_queue.md
Approval Queue Health Audit Script
Zero-token read-only scanner that verifies pending approval queue items include the required hazard marker, proposed action, value, risk notes, approval options, D.E.V. recommendation, and clear safety-boundary wording before J.A.R.V.I.S. relies on the queue as a commander decision surface.
Status: active
Location: /opt/data/profiles/dev/scripts/approval-queue-health-audit.py`; latest artifacts at `/opt/data/agent-bridge/approval-queue/health-audit-latest.json` and `.md
Approval Queue Decision Pack Script
Zero-token local builder that turns pending approval queue items into clean approve/reject/revise decision prompts, exact non-truncated approval-line drafts, recommendations, safety notes, and a Slack-readable founder approval brief so J.A.R.V.I.S. can brief Stufio without rereading the whole queue or accidentally executing approval-gated work. The founder brief orders decisions from the current first-approval ask and proactive decision matrix, strips raw Markdown decoration from Slack-visible `.txt` output, removes duplicate safety-label noise, and preserves approval scope and stop-line wording.
Status: active
Location: /opt/data/profiles/dev/scripts/approval-queue-decision-pack.py`; latest artifacts at `/opt/data/agent-bridge/approval-queue/decision-pack-latest.json`, `.md`, and `/opt/data/agent-bridge/approval-queue/founder-approval-brief-latest.txt
Approval Decision Aging Audit Script
Zero-token local audit that combines pending approval queue dates, the approval decision pack, the current proactive decision matrix, and commander next-moves trend evidence to rank approval decisions by current commander priority, age, and recurrence. It identifies the first exact approval line J.A.R.V.I.S. should ask Stufio for while preserving all safety boundaries, not executing any approval-gated item, preserving timestamped history snapshots for recurring-decision trend review, and writing a clean one-decision founder ask text artifact for Slack-readable commander briefs. The first-approval ask prefers the current decision-matrix rank before older trend history and now strips raw Markdown decoration plus duplicate safety-label noise from Slack-visible `.txt` wording while preserving scope and safety text.
Status: active
Location: /opt/data/profiles/dev/scripts/approval-decision-aging-audit.py`; latest artifacts at `/opt/data/agent-bridge/approval-queue/decision-aging-audit-latest.json`, `.md`, and `/opt/data/agent-bridge/approval-queue/first-approval-ask-latest.txt`; history snapshots under `/opt/data/agent-bridge/approval-queue/decision-aging-history/
Approval First-Ask Trend Audit Script
Zero-token local trend audit that reads approval-decision aging history snapshots, detects whether the same first approval ask is recurring across commander review cycles, and writes a Slack-readable first-ask trend cue so J.A.R.V.I.S. can ask Stufio for the right decision without rereading overwritten latest artifacts. The Slack-visible `.txt` cue now strips raw backticks, asterisks, and pipe-table characters while preserving approval/safety meaning in JSON/Markdown audit artifacts.
Status: active
Location: /opt/data/profiles/dev/scripts/approval-first-ask-trend-audit.py`; latest artifacts at `/opt/data/agent-bridge/approval-queue/first-ask-trend-audit-latest.json`, `.md`, and `/opt/data/agent-bridge/approval-queue/first-ask-trend-latest.txt
Approval Ledger Health Audit Script
Zero-token local audit that scans `/opt/data/agent-bridge/approval_ledger.md`, verifies active approvals preserve source/status/guardrails/reporting evidence, confirms category-level approvals keep concrete-scope and no-secrets boundaries, and writes a commander cue before J.A.R.V.I.S. relies on approval evidence for higher-risk work.
Status: active
Location: /opt/data/profiles/dev/scripts/approval-ledger-health-audit.py`; latest artifacts at `/opt/data/agent-bridge/approval-ledger/health-audit-latest.json` and `.md
Agent Skill/Procedure Log Health Audit Script
Zero-token local audit that scans `/opt/data/agent-bridge/agent_skill_improvement_log.md`, counts skill/procedure improvement entries, detects malformed list-prefix drift, duplicate headings, missing required fields, and weak safety/verification wording so J.A.R.V.I.S. can trust the learning ledger as a commander-readable source of reusable improvements.
Status: active
Location: /opt/data/profiles/dev/scripts/skill-procedure-log-health-audit.py`; latest artifacts at `/opt/data/agent-bridge/agent-skill-improvement/log-health-audit-latest.json` and `.md
Agent Readiness Scorecard
14-category scorecard plus promotion gates, expansion runway readiness, future-lane quickstart confidence, evidence freshness, and confidence-matrix interpretation for deciding whether an agent is ready for expanded responsibility, tools/access, recurring jobs, public/customer-facing workflows, production actions, or higher-risk operations. Includes readiness confidence and drift-state guidance for consistent promotion decisions.
Status: active
Location: /opt/data/agent-bridge/AGENT_READINESS_SCORECARD.md
Agent Readiness Freshness Guard Protocol
Defines mandatory freshness gates (`PASS`/`WATCH`/`ATTENTION`) before any readiness-based expansion or recurring-autonomy decision. Requires fresh readiness index, readiness pulse, scorecard-health evidence, confidence-matrix output, and explicit next safe action and stop path for stale states.
Status: active
Location: /opt/data/agent-bridge/AGENT_READINESS_FRESHNESS_GUARD_PROTOCOL.md
Agent Readiness Confidence Matrix Protocol
Zero-token confidence-matrix protocol and helper that turns scorecard categories, readiness-index status, pulse evidence, and upgrade-card readiness inputs into a one-screen PASS/WATCH/ATTENTION authority-expansion hold/advance cue. Now includes parser hardening so latest readiness pulse snapshots are not falsely reported as missing, markdown field residue is stripped from commander-facing review dates, and each lane's confidence-linked next safe action is visible in the Markdown artifact without requiring J.A.R.V.I.S. to open JSON.
Status: active
Location: /opt/data/agent-bridge/AGENT_READINESS_CONFIDENCE_MATRIX_PROTOCOL.md`; helper at `/opt/data/profiles/dev/scripts/agent-readiness-confidence-matrix.py`; latest artifacts at `/opt/data/agent-bridge/agent-readiness/agent-readiness-confidence-matrix-latest.json` and `.md
Agent Readiness Drift Radar Protocol
Standardizes a compact, artifact-driven review loop that converts readiness index, scorecard health, identity drift, roster/activation hygiene, and pulse status into one-page readiness drift packets with confidence bands and safe next actions before expansion decisions.
Status: active
Location: /opt/data/agent-bridge/AGENT_READINESS_DRIFT_RADAR_PROTOCOL.md
Agent Readiness Snapshot Protocol
One-screen readiness snapshot format that compresses the scorecard into a commander-ready decision cue, confidence band, readiness delta, strongest/weakest areas, verified evidence, stop path, and recommendation so promotion or hold decisions are faster and clearer.
Status: active
Location: /opt/data/agent-bridge/AGENT_READINESS_SNAPSHOT_PROTOCOL.md
Agent Readiness Promotion Review Template
Compact review packet for packaging scorecard results, a decision cue, readiness delta, sandbox drill outcome, blast radius, rollback/stop path, and next safe step into one promotion decision artifact.
Status: active
Location: /opt/data/agent-bridge/AGENT_READINESS_PROMOTION_REVIEW_TEMPLATE.md
Agent Expansion Runway Review Template
One-screen runway gate that combines readiness score, pulse status, promotion review, missing evidence lanes, delegation/depth risk, and explicit next-safe-action/stop path before any authority, tooling, or recurring-autonomy expansion.
Status: active
Location: /opt/data/agent-bridge/AGENT_EXPANSION_RUNWAY_REVIEW_TEMPLATE.md
Agent Readiness Pulse Snapshot Script
Generates a one-screen readiness pulse artifact from `agent-readiness/readiness-index-latest.md` and related scorecard/snapshot/protection templates to support fast, evidence-led hold/internal-autonomy decisions.
Status: active
Location: /opt/data/profiles/dev/scripts/agent-readiness-pulse-snapshot.py
Agent Readiness Index Builder
Zero-token local builder that combines active roster state, future/inactive lanes, scorecard health, resume-matrix coverage, identity drift, future-agent activation gate hygiene, onboarding inheritance, and pending activation decisions into one commander-readable readiness index before J.A.R.V.I.S. expands authority, activates future agents, grants recurring autonomy, or routes access/tool decisions.
Status: active
Location: /opt/data/profiles/dev/scripts/agent-readiness-index-builder.py`; latest artifacts at `/opt/data/agent-bridge/agent-readiness/readiness-index-latest.json` and `.md
Agent Readiness Scorecard Health Audit Script
Zero-token read-only scanner that checks readiness decision surfaces for required categories, score placeholders, snapshot fields, promotion review fields, evidence/approval/stop-path guardrails, and sandbox-drill dependency before J.A.R.V.I.S. relies on readiness materials for agent expansion or authority decisions.
Status: active
Location: /opt/data/profiles/dev/scripts/agent-readiness-scorecard-health-audit.py`; latest artifacts at `/opt/data/agent-bridge/agent-readiness/scorecard-health-audit-latest.json` and `.md
Agent Evidence Packet Protocol
Standard lightweight evidence-packet format for recording checks performed, files/systems inspected or changed, actual verification results, approval/safety notes, and follow-up so future agents can continue work without fabricated context.
Status: active
Location: /opt/data/agent-bridge/AGENT_EVIDENCE_PACKET_PROTOCOL.md
Evidence Packet Health Audit Script
Zero-token local audit that scans `/opt/data/agent-bridge/evidence/` and `/opt/data/agent-bridge/dev-jarvis/` for evidence-packet shape, verification wording, approval/safety wording, and secret-like literal exposure risk so J.A.R.V.I.S. can distinguish usable proof from documentation debt before relying on bridge evidence.
Status: active
Location: /opt/data/profiles/dev/scripts/evidence-packet-health-audit.py`; latest artifacts at `/opt/data/agent-bridge/evidence-packets/health-audit-latest.json` and `.md
Evidence Packet Upgrade Operator Card Protocol
Converts broad evidence-packet health-audit WATCH findings into a small ordered operator card so J.A.R.V.I.S. can standardize the highest-value evidence notes opportunistically without bulk-editing historical evidence or treating documentation debt as live execution authority.
Status: active
Location: /opt/data/agent-bridge/EVIDENCE_PACKET_UPGRADE_OPERATOR_CARD_PROTOCOL.md`; helper at `/opt/data/profiles/dev/scripts/evidence-packet-upgrade-operator-card.py`; latest artifacts at `/opt/data/agent-bridge/evidence-packets/upgrade-operator-card-latest.md`, `.json`, and `.txt
Proactive Loop Health Snapshot Protocol
Compact report backbone for proactive improvement runs that records an explicit `Status: DONE / IN PROGRESS / BLOCKED / APPROVAL NEEDED`, target, improvements made, loop health, active jobs checked, last/next run, delivery target, skill/tooling, drift or duplicate risk, leadership check, Central Command check, verification, next leap, and approval-needed items so Stufio can scan the loop in seconds. Includes a watch/attention escalation ladder requiring local recovery packets before any resume/restart action on flagged channels or provider-limited jobs. Includes an explicit emergency-brake BLOCKED posture when `/opt/data/agent-bridge/team/EMERGENCY-BRAKE-ACTIVE.md` is present.
Status: active
Location: /opt/data/agent-bridge/PROACTIVE_LOOP_HEALTH_SNAPSHOT_PROTOCOL.md
Proactive Loop Run Health & Safe Experiment Protocol
Run-time protocol for combining practical health checks with safe experiment conversion. It requires pre-run health checks, safe local/mock-first experiments, prioritized approval-matrix ordering for high-risk pending items, and clear mapping from bold ideas to reversible actions before any live operational change. The health precheck labels paused/triaged delivery warnings with last-run age context and known active recovery-pending errors with their closure event, so stale held warnings or repaired-but-unverified incidents are not mistaken for fresh unclassified failures. In emergency-brake mode, it explicitly requires non-execution documentation-only posture.
Status: active
Location: /opt/data/agent-bridge/PROACTIVE_LOOP_RUN_HEALTH_AND_EXPERIMENT_PROTOCOL.md
Proactive Improvement Prioritized Decision Matrix Protocol
Zero-token helper protocol + script output workflow for building a one-screen prioritized decision matrix from approval queue risk/aging, proactive precheck posture, approval queue health, and approval-ledger status before attempting any live execution or risky follow-through.
Status: active
Location: /opt/data/agent-bridge/PROACTIVE_LOOP_PRIORITIZED_DECISION_MATRIX_PROTOCOL.md
Proactive Loop Context Resilience Protocol
Escalation protocol for context-length/oversized-output failures in recurring proactive runs. It enforces a continuity packet + split-pass recovery pattern (evidence → synthesis → decision) before any retry, preventing blind repeats or unsafe live retries when context overflows.
Status: active
Location: /opt/data/agent-bridge/PROACTIVE_LOOP_CONTEXT_RESILIENCE_PROTOCOL.md
Proactive Loop Bold Idea to Experiment Protocol
Standard format for turning creative improvement ideas into reversible no-spend, no-live-risk experiment cards with explicit success criteria, evidence path, risk boundary, next safe action, and escalation line for future approval-only steps.
Status: active
Location: /opt/data/agent-bridge/PROACTIVE_LOOP_BOLD_IDEA_TO_EXPERIMENT_PROTOCOL.md
Proactive Report Contract Script
Zero-token local audit that verifies the bridge protocols shaping proactive reports still require the clean status line, `Improvements Made`, J.A.R.V.I.S. leadership check, Central Command check, verification/result evidence, approval-needed queue handling, delegation debt summary/mitigation/validation fields, and Slack-readable formatting reminder before D.E.V. treats automatic reports as commander-ready.
Status: active
Location: /opt/data/profiles/dev/scripts/proactive-report-contract-audit.py`; latest artifacts at `/opt/data/agent-bridge/proactive-loop/report-contract-audit-latest.json` and `.md
D.E.V. Proactive Loop Health Precheck Script
Zero-token deterministic precheck that reads D.E.V. cron list/status, confirms the core proactive loop and safety jobs, checks emergency-brake state, detects drift/warnings, recognizes explicitly triaged paused/retire-candidate jobs such as the Webflow no-new-tabs watchdog, and writes compact bridge artifacts before any risky job/provider/gateway/routing change.
Status: active
Location: /opt/data/profiles/dev/scripts/proactive-loop-health-precheck.py`; latest artifacts at `/opt/data/agent-bridge/proactive-loop/health-precheck-latest.json` and `.md
D.E.V. Proactive Improvement Composite Packet Script
Zero-token read-only helper that runs the proactive-loop health precheck, reads approval-queue/protocol-registry/approval-ledger signals, and folds in the stability scoreboard, domain consultation pulse, proactive WATCH pattern status/count/first-ask alignment, commander next-moves status, Agent Upgrade Card review status/overdue count, N.E.X.U.S. creative-log health/finding count, N.E.X.U.S. reconciliation-preview status/source-integrity/proposal counts, and the exact current first approval ask. Its pending-approval preview now follows the current proactive decision-matrix order instead of raw queue file order, preserving queue titles while keeping `CURRENT` decisions ahead of `SUPERSEDED-REVIEW` items and recording the ordered list/source in JSON and run-delta evidence. It generates a bounded one-screen composite packet and Slack-readable first-read surface, compares commander-safe fields for `CHANGED`, `UNCHANGED`, or `UNCHANGED WATCH`, and preserves full approval/stop boundaries. The commander-cycle builder now regenerates this packet after current cycle/next-moves writes, and the commander freshness index scores it, preventing stale first-read output from sitting beside fresh commander evidence.
Status: active
Location: /opt/data/profiles/dev/scripts/proactive-improvement-composite-packet.py`; latest artifacts at `/opt/data/agent-bridge/proactive-loop/proactive-composite-packet-latest.md`, `.json`, and `.txt`; timestamped history artifacts under `/opt/data/agent-bridge/dev-jarvis/*-proactive-composite-packet.md
Command Center Manual Freshness No-Agent Precheck Script
Zero-token deterministic precheck that scans current Command Center page/API routes, verifies owner/developer manuals mention required routes and safety labels, recognizes Next.js dynamic route bracket syntax, and writes compact bridge audit artifacts before the provider-sensitive LLM-backed manual audit.
Status: active
Location: /opt/data/profiles/dev/scripts/command-center-manual-freshness-precheck.py`; latest artifacts at `/opt/data/agent-bridge/command-center/manual-freshness-precheck-latest.json
Agent Continuity Handoff Packet Protocol
Standard bridge-note packet for continuing work safely after handoffs, corrections, stale-work findings, milestone pauses, future-agent preparation, recurring-job continuation, or RED/YELLOW health-check findings; names owner, recommended next owner, current state, evidence, next safe action, freshness window, approval gates, and stop path.
Status: active
Location: /opt/data/agent-bridge/AGENT_CONTINUITY_HANDOFF_PACKET_PROTOCOL.md
Gateway Restart Safety + Resume-Jobs Standing Rule
Requires J.A.R.V.I.S. to run a safety audit before restarting affected agent gateways, verify Slack-only/Telegram-paused state, avoid duplicate/stale unsafe gateway state, then command affected agents to return to standing roles after verification.
Status: active
Location: /opt/data/agent-bridge/team/2026-06-10-gateway-restart-safety-resume-rule.md
N.E.X.U.S. Innovation DNA Protocol
Standing self-start protocol requiring N.E.X.U.S. to proactively generate, log, and surface Cloudland innovation opportunities; includes scheduled pulse automation, creative-log requirements, D.E.V. support request, safety/approval gates, and a Pilot-to-Production Confidence Gate so ideas remain clearly separated from approved implementation work.
Status: active
Location: /opt/data/agent-bridge/NEXUS_INNOVATION_DNA_PROTOCOL.md
N.E.X.U.S. Creative Log Protocol
Defines how N.E.X.U.S. records ideas in markdown and CSV with required fields, status/implementation vocabularies, recall behavior, and safety boundaries. Now links the D.E.V. zero-token creative-log health audit so J.A.R.V.I.S. can check parity, duplicate IDs, required fields, and vocabulary drift before relying on the log as a commander decision surface or future Command Center opportunity-card source.
Status: active
Location: /opt/data/agent-bridge/nexus-jarvis/NEXUS_CREATIVE_LOG_PROTOCOL.md
N.E.X.U.S. Creative Log Health, Reconciliation Preview, and Review Operator Card Scripts
Zero-token tooling that audits markdown/CSV ID parity, duplicate IDs, required fields, approved vocabularies, title mismatches, and near-duplicate titles, then generates non-mutating append-ready review artifacts and conservative vocabulary proposals when ATTENTION exists. The review operator card converts the large proposal set into five ordered, HOLD-by-default batches, puts unresolved authority semantics first, and gives N.E.X.U.S./J.A.R.V.I.S. explicit acceptance and stop criteria without editing either source. All three helpers run in the recurring commander-cycle chain; artifacts are freshness-scored and the Slack-readable card is format-audited.
Status: active
Location: health audit `/opt/data/profiles/dev/scripts/nexus-creative-log-health-audit.py`; reconciliation preview `/opt/data/profiles/dev/scripts/nexus-creative-log-reconciliation-preview.py`; review card `/opt/data/profiles/dev/scripts/nexus-reconciliation-review-operator-card.py`; latest artifacts under `/opt/data/agent-bridge/nexus-jarvis/audit/
Gateway Restart Back-Online Acknowledgement Addendum
Addendum requiring each affected agent to acknowledge it is back online in the active operations channel after gateway restart verification, then return to standing duties; J.A.R.V.I.S. also resumes commander duties.
Status: active
Location: /opt/data/agent-bridge/team/2026-06-10-gateway-restart-back-online-addendum.md
Agent Standing Duties Resume Matrix
Commander-readable matrix that maps each active agent to its standing lane, first safe post-restart/recalibration action, bridge/evidence path, and approval-sensitive stop line. Future agents must add a row or linked resume card before being treated as ACTIVE. Now links the zero-token roster/resume coverage audit and standing-duties resume pack generator so J.A.R.V.I.S. can both catch missing coverage and generate precise per-agent resume instructions after safety verification.
Status: active
Location: /opt/data/agent-bridge/AGENT_STANDING_DUTIES_RESUME_MATRIX.md
Agent Roster Resume Matrix Coverage Audit Script
Zero-token read-only audit that compares `/opt/data/agent-bridge/ACTIVE_AGENT_ROSTER.md` against `/opt/data/agent-bridge/AGENT_STANDING_DUTIES_RESUME_MATRIX.md`, confirms every ACTIVE agent has a resume-matrix row with a first safe action and approval stop line, and flags missing bridge evidence surfaces as WATCH instead of live failure.
Status: active
Location: /opt/data/profiles/dev/scripts/agent-roster-resume-matrix-audit.py`; latest artifacts at `/opt/data/agent-bridge/agent-roster/resume-matrix-coverage-latest.json` and `.md
Agent Roster Identity Drift Audit Script
Zero-token read-only audit that detects when the active-agent roster says an agent is inactive/future/concept while local profile or bridge evidence suggests operational setup exists. It gives J.A.R.V.I.S. an ATTENTION/WATCH/PASS commander cue before routing, activation, resume-matrix, or future-agent promotion decisions.
Status: active
Location: /opt/data/profiles/dev/scripts/agent-roster-identity-drift-audit.py`; latest artifacts at `/opt/data/agent-bridge/agent-roster/identity-drift-audit-latest.json` and `.md
Standing Duties Resume Pack Generator
Zero-token read-only generator that converts the active roster and standing-duties resume matrix into commander-ready per-agent resume commands, first safe actions, evidence surfaces, and approval stop lines for J.A.R.V.I.S. after gateway restart, recalibration, Slack recovery, token-conservation wake-up, or future-agent activation checks.
Status: active
Location: /opt/data/profiles/dev/scripts/standing-duties-resume-pack-generator.py`; latest artifacts at `/opt/data/agent-bridge/agent-roster/standing-duties-resume-pack-latest.json` and `.md
Gateway Restart Log Health Audit Script
Zero-token read-only audit that scans `/opt/data/agent-bridge/team/restart-recovery-log.md` for required restart evidence signals: safety audit, profile scope, Slack-only/Telegram-paused posture, no-secrets wording, credential/access scope clarity, online verification, back-online acknowledgement, and standing-duty resume evidence. It gives J.A.R.V.I.S. a PASS/WATCH/ATTENTION cue before relying on restart notes as commander proof.
Status: active
Location: /opt/data/profiles/dev/scripts/gateway-restart-log-health-audit.py`; latest artifacts at `/opt/data/agent-bridge/gateway-restarts/restart-log-health-latest.json` and `.md
Gateway Restart Supplemental Evidence Plan Protocol
Zero-token documentation-only planner that converts gateway restart log `WATCH`/`ATTENTION` audit findings into a prioritized supplemental evidence plan, naming missing restart signals, existing supplemental packet paths, packet fields, next safe action, and stop path before J.A.R.V.I.S. relies on restart/resume evidence.
Status: active
Location: /opt/data/agent-bridge/GATEWAY_RESTART_SUPPLEMENTAL_EVIDENCE_PLAN_PROTOCOL.md`; helper at `/opt/data/profiles/dev/scripts/gateway-restart-supplemental-evidence-planner.py`; latest artifacts at `/opt/data/agent-bridge/gateway-restarts/supplemental-evidence-plan-latest.md`, `.json`, and `.txt
Future Agent Activation Gate Audit Script
Zero-token read-only audit that checks the Future Agent Onboarding Checklist for required acceptance-gate terms, confirms ACTIVE roster agents have resume-matrix coverage, distinguishes legacy active agents from new promotions, and requires explicit activation acceptance packets before future agents are listed ACTIVE or expanded into live/autonomous authority.
Status: active
Location: /opt/data/profiles/dev/scripts/future-agent-activation-gate-audit.py`; latest artifacts at `/opt/data/agent-bridge/future-agent/activation-gate-audit-latest.json` and `.md
Future Agent Onboarding Inheritance Audit Script
Zero-token local audit that checks whether the Future Agent Onboarding Checklist explicitly inherits the required safety, memory, recalibration, post-action reporting, approval queue/ledger, readiness scorecard, sandbox drill, activation gate, active roster, standing-duty resume, protocol registry, skill-log, and continuity handoff surfaces before J.A.R.V.I.S. treats the checklist as an activation decision surface.
Status: active
Location: /opt/data/profiles/dev/scripts/future-agent-onboarding-inheritance-audit.py`; latest artifacts at `/opt/data/agent-bridge/future-agent/onboarding-inheritance-audit-latest.json` and `.md
Agent Slack Gateway Watchdog Standing Rule
Standing approval for J.A.R.V.I.S. to scan all current/future specialist agent Slack gateway statuses every 15 minutes and automatically attempt safe recovery for any discovered specialist profile gateway that is down, while keeping Telegram paused and preserving high-risk approval gates.
Status: active
Location: /opt/data/agent-bridge/team/2026-06-10-agent-slack-gateway-watchdog-rule.md
System Overview Live Health Panel
Upgrades the home-page System Overview panel from a hardcoded static "98%" donut to a live async panel reading real action store data. Health signal (NOMINAL/ATTENTION/ACTION REQUIRED) and row breakdown (Active/Blocked/Completed/Total logged) are computed from the live action log. The donut arc color reflects real health. J.A.R.V.I.S. now has two independent live health surfaces on the home page (Action Health Badge + System Overview Panel). 20 new regression tests.
Status: active
Location: src/components/DashboardShell.tsx` (`SystemOverviewPanel`), `test/systemOverviewLiveHealth.test.mjs` (20 tests), evidence `2026-06-13T161608Z-system-overview-live-health.md
Commander Action Intelligence Panel — Action Store Summary API
Adds a `summarizeActionRecords()` function to the action store providing health signal (green/amber/red), status/agent/priority distributions, approval-needed spotlight, and blocked-item spotlight — all computed from the local action log. The `/actions` page renders a Commander Action Intelligence Panel so J.A.R.V.I.S. can read queue state at a glance. The `GET /api/actions/summary` endpoint returns the same `ActionSummary` JSON directly so any Command Center client (dashboard home, future native app) can display a live action-health badge with a single cheap GET call. The dashboard home now embeds a live `ActionHealthBadge` inside the Tasks panel showing real queue health (NOMINAL / ATTENTION / ACTION REQUIRED) with active count, status breakdown, and conditional blocked/approval-needed callouts linked to `/actions`. The 409 dedup response includes `existingTitle` and `existingAgent` for richer conflict context. 32 new tests added across two runs; full suite at 282/282.
Status: active
Location: src/lib/actionStore.js` (`summarizeActionRecords`), `src/app/actions/page.tsx` (Commander panel), `src/app/api/actions/route.ts` (enriched 409), `src/app/api/actions/summary/route.ts` (dedicated summary endpoint), `src/components/DashboardShell.tsx` (`ActionHealthBadge` inside TasksPanel)
Recurring Job Triage Ledger Protocol
Commander-readable classification layer for recurring jobs so J.A.R.V.I.S. can distinguish HEALTHY, WATCH, HOLD / TRIAGE, APPROVAL NEEDED, and RETIRE CANDIDATE jobs with a clear commander cue, approval sensitivity, evidence source, and next safe action. Includes an evidence-freshness rule requiring UTC update stamps and live cron/precheck evidence so stale next-run/last-run values are not carried forward as current state. The zero-token ledger health audit checks freshness, required fields, classification/cue vocabulary, evidence wording, and HOLD / TRIAGE anti-rerun language before J.A.R.V.I.S. relies on the ledger. The zero-token refresh helper rewrites the live ledger from current Hermes cron output while sanitizing raw provider payloads/request IDs and now preserves the Webflow no-new-tabs watchdog as a paused RETIRE CANDIDATE instead of an unclassified WATCH lane.
Status: active
Location: /opt/data/agent-bridge/RECURRING_JOB_TRIAGE_LEDGER_PROTOCOL.md`; live ledger `/opt/data/agent-bridge/recurring_job_triage_ledger.md`; health audit script `/opt/data/profiles/dev/scripts/recurring-job-triage-ledger-health-audit.py`; refresh helper `/opt/data/profiles/dev/scripts/recurring-job-triage-ledger-refresh.py`; latest health artifacts `/opt/data/agent-bridge/recurring-jobs/triage-ledger-health-audit-latest.json` and `.md`; latest refresh artifacts `/opt/data/agent-bridge/recurring-jobs/triage-ledger-refresh-latest.json` and `.md
D.E.V. Task Intake Reconciliation Protocol
Read-only reconciliation procedure for stale `inprog-*` task markers so J.A.R.V.I.S. can distinguish evidence-backed completed work from true active or blocked D.E.V. tasks without deleting files or guessing from old acknowledgements. Links the zero-token task intake reconciliation audit helper so stale markers get a fresh commander-readable classification before cleanup, re-dispatch, or closure decisions. Now also requires D.E.V. task-watcher Slack updates to use clean `D.E.V. —`, uppercase-label formatting with no raw markdown/code ticks in visible Slack text.
Status: active
Location: /opt/data/agent-bridge/DEV_TASK_INTAKE_RECONCILIATION_PROTOCOL.md
D.E.V. Task Intake Reconciliation Audit Script
Zero-token read-only scanner that checks stale `inprog-*.json` task markers against ack files and bridge completion evidence, classifies them as evidence-backed, ack-only, no-ack, or approval-sensitive hold, and writes a commander cue before J.A.R.V.I.S. treats stale D.E.V. task intake as active, blocked, complete, or ready for cleanup.
Status: active
Location: /opt/data/profiles/dev/scripts/dev-task-intake-reconciliation-audit.py`; latest artifacts at `/opt/data/agent-bridge/dev-tasks/audit/intake-reconciliation-latest.json` and `.md
Rate Limit Emergency Brake Protocol
Tiered rate-limit alert and emergency brake system. A no-agent cron (job `ff9536fa41d0`, every 15m) runs the watchdog script using zero LLM tokens. The script checks actual runtime 429 signals in gateway logs and cron session dumps (NOT documentation files — false-positive bug fixed 2026-06-13). Tiered alerts: $5 WARN, $2 ALERT, $1 BRAKE. At BRAKE threshold: writes brake file, pauses scoped D.E.V. cron jobs, notifies Stufio. Resume requires Stufio saying "Resume all systems". Anti-loop rule: provider failures become pause/triage/continuity handoff, not repeated LLM retries.
Status: active
Location: /opt/data/agent-bridge/dev-jarvis/RATE-LIMIT-EMERGENCY-BRAKE-PROTOCOL.md`; watchdog script `/opt/data/profiles/dev/scripts/rate-limit-watchdog.py`; brake file `/opt/data/agent-bridge/team/EMERGENCY-BRAKE-ACTIVE.md`; state file `/opt/data/agent-bridge/team/rate-limit-watchdog-state.json
Recurring Provider Failure Recovery Matrix
Commander recovery matrix for recurring job failures caused by HTTP 429, quota/cap/low-credit pressure, Slack delivery failures, high-frequency LLM loop risk, and stale paused jobs. Defines HEALTHY, WATCH, HOLD / TRIAGE, APPROVAL NEEDED, and RETIRE CANDIDATE handling, plus no-agent replacement and recovery-packet formats so J.A.R.V.I.S. can preserve useful automation without retry loops or surprise spend.
Status: active
Location: /opt/data/agent-bridge/RECURRING_PROVIDER_FAILURE_RECOVERY_MATRIX.md
J.A.R.V.I.S. Command Center — Zero-Failure Test Suite (346/346)
All 5 remaining test failures fixed in 2026-06-13 loop-2. JarvisCoreDemo upgraded: real mic-level monitoring via getUserMedia + requestAnimationFrame, speaking state via setSpeaking/useState, speaking-state-demo-toggle testid with aria-pressed, core-voice-level-meter testid, Voice/Idle mode labels. VoiceProviderAdapters refactored: fetch() network call removed from adapter (config/routing only); actual HTTP call belongs in route handler. Full test suite: 346/346 passing; build: clean.
Status: active
Location: src/components/JarvisCoreDemo.tsx`, `src/lib/voiceProviderAdapters.js`, `test/` directory (346 tests)
Proactive Loop Stability Scoreboard Protocol
Creates a one-screen stability gate for each proactive cycle by combining precheck status, decision-matrix risk, approval-queue volume, and the current first-approval ask. It emits an explicit execution recommendation, JSON fields for downstream commander-cycle builders, and watch-only allowed updates when WATCH/attention risk remains active, preventing blind retries and unsafe live follow-through.
Status: active
Location: /opt/data/agent-bridge/PROACTIVE_LOOP_STABILITY_SCOREBOARD_PROTOCOL.md`; script at `/opt/data/profiles/dev/scripts/proactive-loop-stability-scoreboard.py`; artifacts `/opt/data/agent-bridge/proactive-loop/stability-scoreboard-latest.md` and `.json
Proactive Loop Domain Consultation Pulse Protocol
Adds a required one-screen per-run consultation hygiene check for proactive-loop work that touches domain-owned lanes. Standardizes consultation state (`performed` / `blocked` / `deferred` / `not applicable`), evidence path, ownership traceability, and next safe action. The zero-token helper now writes latest and timestamped history artifacts so J.A.R.V.I.S. can see when cron runs used bridge evidence only and must not drift into live Slack/gateway/routing/provider/job changes.
Status: active
Location: /opt/data/agent-bridge/PROACTIVE_LOOP_DOMAIN_CONSULTATION_PULSE_PROTOCOL.md`; helper script at `/opt/data/profiles/dev/scripts/proactive-loop-domain-consultation-pulse.py`; latest artifacts at `/opt/data/agent-bridge/proactive-loop/domain-consultation-pulse-latest.md` and `.json