fix(termux-multi-agent): add fallback rich UI classes in dashboard.py

.agents/skills/adaptive-wait/SKILL.md

@@ -1,29 +1,125 @@
 ---
 name: adaptive-wait
-description: Adaptive WAIT stage for agentic GitHub ops. After any dispatch, commit, rebase, or merge request — re-check jobs/steps/logs/artifacts, do concurrent non-conflicting work, classify stalls, and only promote when dual gates + task outcome are verified. Triggers when waiting on CI, after PR open/update, during /continue cycles, or when operator says wait adaptively. Cross-session SSOT — always load from master, never chat-only.
+description: Adaptive WAIT for agentic GitHub operations. Re-fetch authoritative state, detect progress or stalls, keep useful disjoint work moving, and promote only from current-SHA evidence.
 ---
 
-# Skill: adaptive-wait
+# Adaptive WAIT
 
-**Owner:** ArchW1z / Grok Administrator  
-**Complements:** `evidence-led-monorepo-ops`, `adaptive-feedback-cycle`, `production-reconciliation`, `review-loop`
+Use this skill whenever work is asynchronous: GitHub Actions, reviews, deployments, provider jobs, external agents, or any operation whose next action depends on new state.
 
-While one PR's checks run: preserve immutable IDs; work on a disjoint path; do not idle.
+## Control objective
 
-Promote only when dual gates success, extract-clean scope, and task outcome verified.
+Waiting is an evidence-collection phase, not an idle timer. HOLD is not an idle parking state.
 
-## Failure / stall classes (do not soft-pedal)
+The controller must answer, on every cycle:
 
-| Class | Severity | Notes |
-|-------|----------|-------|
-| **comment-storm** | **FAILURE** | issue_comment / bot / ledger fan-out that cancels useful jobs. Mitigate: concurrency by event_name, `cancel-in-progress: false` on ledgers. Landed #603 on `dc30bf83`. |
-| dual-gate red | FAILURE | Block promote |
-| update-branch-conflict | STALL | Extract-later; do not force dirty |
-| dirty-behind-master | STALL | Rebase/extract from live master |
-| extra-red | FAILURE (non-gate) | Fix root cause |
+1. What immutable state am I watching?
+2. What evidence changed since the last observation?
+3. Can that evidence change the next action?
+4. If not, should I back off, do disjoint work, or stop?
+5. What authoritative re-fetch will prove the next state?
 
-HEAD after #603: `dc30bf83fc17b394510f99328f7a081b6a64f28c`. #601 ML dirty/behind — WAIT extract.
+Continuous Fully Automated Agentic Development: agents auto-promote when dual-gate + task outcome are verified on the candidate SHA. Sovereignty cockpit = ArchWiz + chat (goals/constraints, not a synchronous merge checkpoint).
 
-Session 2026-09-18T17:03Z: merged #603; CodeRabbit `queue: max` rejected (invalid GHA concurrency key).
+## State machine
 
-BIUDL. Agent-Identity: Grok (Administrator)
+| State | Entry evidence | Action |
+|---|---|---|
+| ACTIVE | run/job/review/artifact state is changing or useful work is still executing | re-fetch at an adaptive cadence; do not duplicate work |
+| QUIET | no meaningful delta, but the operation remains live | increase the wait interval; work on a disjoint bounded task |
+| STALLED | repeated identical state, no progress, deadlock, expired deadline, or repeated non-informative retry | stop blind retries; classify and change the experiment/approach |
+| TERMINAL | authoritative success/failure/cancellation or task outcome is verified | validate current SHA and close the cycle |
+| BLOCKED | required authority, credential, input, or external capacity is missing | record the blocker; request only the missing authority/input |
+
+Never treat elapsed time, HTTP 200, a successful dispatch, or a green individual job as task completion.
+
+## Adaptive cadence
+
+Choose the next observation from evidence, not a universal polling constant:
+
+- Immediate re-fetch: after a material commit, workflow dispatch, rerun, steering action, review response, or newly reported failure.
+- Short wait: while a run/job is actively producing new steps, logs, artifacts, or review events.
+- Backoff: when repeated observations contain no material delta and the authoritative state remains live.
+- Stop retrying: when the same failure repeats without a changed input, environment, or hypothesis.
+- One-shot: when no new feedback can change the next action.
+- Deadline stop: when the platform/task deadline is authoritative; do not invent a second application-level timeout.
+
+Record why the cadence changed.
+
+## Mandatory re-fetch contract
+
+After every material commit or steering action, re-fetch: SHA -> workflow/run -> jobs -> steps/logs/artifacts -> reviews/comments -> resulting SHA/status.
+
+Bind every conclusion to the SHA that produced its evidence. A newer SHA invalidates approval evidence from an older head unless the evidence is explicitly historical.
+
+## Disjoint-work rule
+
+While useful work is running, continue only with work that cannot mutate or invalidate the watched operation. Good disjoint work includes deterministic documentation or inventory improvements, bounded repository reconnaissance, non-invasive tests, relationship-graph queries, and review/evidence ingestion.
+
+Do not create competing writes, duplicate workflow dispatches, or overlapping branches merely to stay busy.
+
+## Retry / steering rule
+
+Retry only when the new attempt has an evidence-backed reason to differ: changed input or SHA; transient infrastructure/provider failure; recovered capacity/quota; corrected workflow/reference; or a new review finding or authoritative instruction.
+
+Preserve every attempt. Never rewrite history to hide a failed wait cycle.
+
+## Evidence receipt
+
+Each cycle should be representable by a compact receipt:
+
+    observed_at: <UTC>
+    watched:
+      repo: <owner/name>
+      ref: <branch/tag/SHA>
+      sha: <immutable SHA>
+    state: ACTIVE|QUIET|STALLED|TERMINAL|BLOCKED
+    evidence:
+      runs: [<run ids>]
+      jobs: [<job ids>]
+      reviews: [<stable ids>]
+      artifacts: [<stable ids>]
+    delta: <what changed since prior observation>
+    decision: WAIT|BACKOFF|STEER|RETRY|STOP|PROMOTE
+    reason: <evidence-backed reason>
+    next_check: <authoritative condition>
+    outcome: PASS|FAIL|UNKNOWN
+
+Never store secrets, authorization headers, raw discussion bodies, or private session state in the receipt.
+
+## Promotion boundary
+
+Promotion requires: current immutable SHA; current-base relationship verified; relevant validation evidence terminal; requested task outcome verified; no unresolved actionable finding; and evidence attributable to the candidate SHA.
+
+Dual-gate status is necessary where the repository defines it, but it is not a substitute for task-outcome verification.
+
+Repository dual-gate: `hygiene + portability gate` + `agentic termux smoke`. Vercel rate-limit is non-gate (#772). Copilot / CodeRabbit / Devin are advisory only. `mergeable_state=dirty` or a GitHub merge conflict is a hard block even when dual-gate is green — rebase/re-extract onto live tip.
+
+**Mega-merge (Operator 2026-09-24):** Allowed when dual-gate + mergeable. CodeRabbit ~100-file limit is advisory review capacity, not a promote ban. Size alone does not block.
+
+## Terminal states
+
+- success — requested outcome verified;
+- clean-no-op — no change required;
+- blocked — new authority/input is required;
+- approval-required — policy requires escalation;
+- exhausted — authoritative task/platform limit reached;
+- stagnated — repeated cycles produced no measurable progress.
+
+Never report exhausted, stalled, or blocked as success.
+
+## Closeout
+
+Record what was proven, what remains unproven, the watched SHA, the final authoritative evidence, and the next disjoint action or terminal state.
+
+This skill is the wait/controller layer. adaptive-feedback-cycle owns broader continuous learning; production-reconciliation owns ref alignment; evidence-envelope owns normalized observation fields.
+
+## Session 2026-09-28 17:15 PDT
+
+- Live master `8daeeb72d71ecdefa2f9cde6698426131117e474`.
+- Dual-gate PASS: repo-gate 36497249095 / termux-smoke 36497249115.
+- #904 merged. Watching #899 rebase onto live tip.
+- Instant-fail path-unfiltered workflows are noise.
+- #903 HOLD. Do not merge #892/#894. Do not pulse #175. #184 names-only.
+
+Agent-Identity: Grok (Administrator)

.agents/skills/adaptive-wait/references/BOARD-VS-LEDGER.md

@@ -0,0 +1,7 @@
+# Adaptive wait + board
+
+WAIT collects evidence on a product SHA. It does not create a new pulse PR.
+Read generated lane-matrix-status. Close leftover session pulses as SUPERSEDE.
+Do not comment #175 as a heartbeat.
+
+Session 2026-09-24 12:33 PDT. Agent-Identity: Grok (Administrator)

.agents/skills/context-relationship-graph/SKILL.md

@@ -48,6 +48,33 @@ Do not persist discussion bodies, session stores, browser profiles, credentials,
 
 Every PR/issue is an observation candidate. Not every PR needs a Markdown receipt. Receipts are human-readable projections of notable measurements or promotions.
 
+## Temporal evidence layer
+
+The historical backfill now produces a second, append-only evidence layer under:
+
+`workspace/llm_map/context_relationships/temporal/`
+
+Each observation receives a deterministic `snapshot_id` bound to repository/ref, source SHA, observation time, and page bounds. Snapshot manifests retain:
+
+- source SHA/ref and observation timestamp;
+- history start/continuation pages;
+- explicit coverage state;
+- node/edge counts and content hashes;
+- previous snapshot identity;
+- structural delta counts.
+
+Snapshots are immutable. Re-running the same observation may reuse the existing snapshot but must never overwrite its contents.
+
+The temporal layer distinguishes:
+
+`source → relationship → observation → snapshot → lineage`
+
+A later observation may add, remove, change, or reclassify a relationship. This is a delta, not permission to rewrite the prior evidence.
+
+Use `temporal-current.json` for the latest materialized observation and `temporal/lineage.jsonl` for the append-only observation chain. The current graph remains the L1 relationship view; temporal snapshots are the L2 evidence/history layer consumed by SHE, longitudinal analysis, and future context reconstruction.
+
+Coverage is complete only when `history_window.next_start_page == null`. Node/edge growth alone never establishes historical completeness.
+
 ## Repository commands
 
 Run from the repository root. Query an exact issue, a specific permalink, or a bounded file-review timeline:

.agents/skills/evidence-led-monorepo-ops/SKILL.md

@@ -1,42 +1,159 @@
 ---
 name: evidence-led-monorepo-ops
-description: Continuous evidence-led admin ops on timerloggedout-spec/termux-monorepo (and similar agentic monorepos). Triggers on priority matrix, master gates, SHE progress, Manus/provider RE, dirty PR triage, Actions hygiene, or when the operator says continue, BIUDL, maximize actions, or /continue. Use for live state pulls, dispositions, small-green extracts, adaptive WAIT, and iterative process improvement documented as skills. Load this skill in every admin session.
+description: Continuous evidence-led admin operations for timerloggedout-spec/termux-monorepo. Reconstruct current state, bind actions to immutable SHAs, preserve provenance, and never confuse activity with verified outcome.
 ---
 
-# Skill: evidence-led-monorepo-ops
+# Evidence-Led Monorepo Operations
 
-**Owner:** ArchW1z / Grok Administrator continuous admin on timerloggedout-spec/termux-monorepo.
+Load this skill for every repository-admin session.
 
-**Canonical paths (keep in sync):**
-- `.agents/skills/evidence-led-monorepo-ops/SKILL.md` ← **agent load path**
-- `docs/ops/skills/evidence-led-monorepo-ops/SKILL.md` ← human/docs mirror
-- `docs/ops/SKILLS-INVENTORY.md` ← full skill table + adaptive WAIT
+## Operating contract
 
-**Primary agent entry:** `CLAUDE.md`.
+RECON -> CLASSIFY -> PLAN -> ACT -> WAIT -> VALIDATE -> RE-FETCH -> RECORD -> REPEAT
 
-## Posture
+The goal is a verified repository outcome, not a green-looking activity stream.
 
-- Evidence over anecdote. Extract-only. Dual-gate before merge.
-- Extra-red ≠ dual-gate. Behind-master dual-gate green ≠ auto-merge.
-- **comment-storm is a FAILURE**, not skippable noise. Mitigate with concurrency groups + `cancel-in-progress: false` on SHA-bound ledgers. Do not treat cancelled ledger runs as green.
-- Identity: `Agent-Identity: Grok (Administrator)`.
-- GitHub MCP write works as `timerloggedout-spec` even when local sandbox has no OPERATOR PAT / `gh`.
+## 1. Reconstruct current state
 
-## Current production anchors (2026-09-18T17:03Z UTC)
+Before a consequential action, resolve:
 
-| Item | State |
-|------|-------|
-| Master HEAD | `dc30bf83fc17b394510f99328f7a081b6a64f28c` (#603 squash) |
-| Just landed | #603 comment-storm ledger fix. Prior: #602 catalog MD fingerprint; #600 skills SSOT. |
-| Observe | #583 Grafana MCP; #584 MVT budget. Jules #597/#598. |
-| Extra-red HOLD | #601 ML extract dirty/behind (`c082f533` base vs live `dc30bf83`). #549/#432 HOLD extract-later. |
-| HOLD mega | #523 #527 #545 #543 #455 #485 #483 #481 |
-| Draft | #578 accounting/bidding schema pilot |
+- live master SHA;
+- target branch/ref and immutable head SHA;
+- merge-base and ahead/behind relationship;
+- changed paths;
+- open PR/issue context;
+- current workflow runs, jobs, steps, artifacts, reviews, and comments relevant to the current SHA;
+- applicable skill, governance, lane, and proposal SSOTs.
 
-## Extract recipe
+Prefer exact file, symbol, PR, issue, label, scope, or permalink roots. Keep verified relationships separate from heuristic candidates.
 
-Create branch from current master. Do not force-update dirty branches. Do not wholesale-merge HOLD mega or #601/#549/#432.
+## 2. Evidence hierarchy
 
-CodeRabbit `queue: max` on GHA concurrency is **not a valid key** (only `group` + `cancel-in-progress`). Do not apply.
+Prefer evidence in this order:
 
-BIUDL. Agent-Identity: Grok (Administrator)
+1. current-SHA repository diff and contracts;
+2. current-SHA validation/test results;
+3. current-SHA workflow/job/step/artifact evidence;
+4. substantive current-SHA review findings;
+5. provenance and task/issue lineage;
+6. historical evidence from superseded SHAs;
+7. size, age, comment count, or activity volume as **context only** (not promote blockers).
+
+A workflow success proves that workflow result. It does not prove the requested task outcome.
+
+## 3. Evidence identity
+
+Every material observation should bind, where available, to repository/ref; immutable commit SHA; base SHA/merge-base; workflow/run/attempt/job/step IDs; artifact or receipt IDs; review/comment IDs; observed/event timestamps; agent/provider/model identity with confidence; experiment/cohort ID when applicable; status/outcome; provenance/confidence; and superseded predecessor.
+
+Never infer missing IDs, timestamps, counts, authorship, or causal relationships.
+
+## 4. State classification
+
+Keep these states distinct:
+
+PASS | FAIL | UNKNOWN | WARNING | SKIPPED | STALLED
+
+Apply them separately to dispatch, execution, provider availability, quota/capacity, correctness, tests, integration, review, deployment, and task outcome.
+
+A provider outage, rate limit, skipped reviewer, stale branch, and correctness failure are different observations.
+
+## 5. Adaptive WAIT integration
+
+During asynchronous work, use .agents/skills/adaptive-wait/SKILL.md.
+
+- Re-fetch immediately after material commits, reruns, steering, or newly reported failures.
+- Back off only when authoritative state is live and repeated observations show no material delta.
+- Retry only when the next attempt differs for an evidence-backed reason.
+- Continue disjoint work while useful evidence is accumulating.
+- Stop on authoritative terminal state, stagnation, or a missing authority/input.
+- Bind conclusions to the SHA that produced the evidence.
+
+HOLD and OBSERVE are not idle parking states. WAIT collects evidence until promote conditions hold, then promote.
+
+## 6. Change discipline
+
+For repository mutations:
+
+1. identify a change that advances the requested outcome (slice **or** mega — both valid);
+2. re-read the current file/ref before writing;
+3. make the change;
+4. run the smallest relevant deterministic validation;
+5. commit with a specific message;
+6. re-fetch the resulting SHA and checks;
+7. preserve failed attempts and superseded evidence.
+
+Never force-push, reset, delete evidence, or silently overwrite another active state.
+
+## 7. Promotion
+
+Promotion requires current-SHA evidence for base alignment/reconciliation; relevant tests and invariants; review/finding disposition; integration/task outcome; and repository-defined dual gates.
+
+COMMITTED, EXECUTED, VALIDATED, and PROMOTED are independent states.
+
+Do not promote because a branch is old, a PR is green on an older SHA, a reviewer is silent, or a provider returned HTTP 200.
+
+Repository dual-gate: hygiene/portability + agentic termux smoke. Vercel is non-gate (#772). Copilot / CodeRabbit / Devin are advisory only.
+
+**Mega-merge policy (Operator 2026-09-24):** Mega-merge is **allowed** when dual-gate SUCCESS + mergeable on the candidate SHA. Collaborators hold large context windows. CodeRabbit’s ~100-file limit is **advisory review capacity**, not a hard promote ban. File count alone must not refuse a dual-gate-green PR. `mergeable_state=dirty` or merge conflict remains a hard block until rebased.
+
+## 8. Historical continuity
+
+Historical collection is bounded and resumable. Record page/window bounds, continuation state, collection timestamp, source/ref, parser/API failures, exclusions, and coverage. Never call a partial page complete history.
+
+A failed or superseded attempt remains evidence. Correct forward with a successor change.
+
+## 9. Write boundaries
+
+Read-only reconnaissance may inspect repository/GitHub state. Mutating operations require the applicable explicit authorization and governance tier.
+
+Never persist secrets, authorization headers, tokens, browser/session stores, private discussion bodies, or generated credentials in the repository. Issue #184 is names-only.
+
+## 10. Closeout receipt
+
+Use a compact receipt shape:
+
+    operation: <bounded operation>
+    repo: <owner/name>
+    base_sha: <sha>
+    head_sha: <sha>
+    observed_at: <UTC>
+    evidence:
+      checks: [<ids>]
+      runs: [<ids>]
+      reviews: [<ids>]
+      artifacts: [<ids>]
+    state: PASS|FAIL|UNKNOWN|WARNING|SKIPPED|STALLED
+    outcome: PASS|FAIL|UNKNOWN
+    provenance: <source/actor confidence>
+    decision: KEEP|RETRY|STEER|HOLD|PROMOTE|STOP
+    reason: <evidence-backed reason>
+    remaining: <unproven work>
+
+Receipts are projections; the longitudinal GitHub/evidence corpus remains the source of historical truth.
+
+## Related operational skills
+
+- adaptive-wait — asynchronous control and cadence;
+- adaptive-feedback-cycle — continuous feedback and historical evaluation;
+- review-loop — review ingestion and repeated validation;
+- context-relationship-graph — metadata-only relationship evidence;
+- production-reconciliation — ref/base reconciliation;
+- action-effectiveness-ledger — action-to-outcome measurement;
+- evidence-envelope / evidence-provenance — normalized provenance;
+- workflow-orchestration — modular Actions coordination;
+- evolutionary-replay — bounded replay of realized discovery history.
+
+This skill owns repository-admin evidence discipline; it does not replace those specialist lanes.
+
+## Session 2026-09-28 17:15 PDT
+
+- Live master tip: `8daeeb72d71ecdefa2f9cde6698426131117e474` (catalog/mmdc bot refresh after #904).
+- Dual-gate PASS on that SHA: repo-gate 36497249095 + termux-smoke 36497249115.
+- #904 squash `4efef68d` dual-gate PASS on master: 36497147812 / 36497147650.
+- Candidate evidence for #904 pre-merge: repo-gate 36497057302 + termux-smoke 36497057157 on `4df79c30`.
+- Tunnel scheduled fail 36496185804 / 109176191755 classified empty-URL skip, closed by #904.
+- Instant-fail path-unfiltered workflows on feature branches are noise.
+- #903 HOLD (empty-diff / ledger). #899 rebase in flight. Do not merge stale #892/#894.
+- Do not pulse #175. #184 names-only. Vercel combined-status is #772 non-gate.
+
+Agent-Identity: Grok (Administrator)