Run any Skill in Manus with one click

reflect-verify

Stars1

Forks1

UpdatedJune 28, 2026 at 20:38

Verify Reflect end-to-end health, especially drain, ingest/consolidate maintenance, launchd timers, statusline error surfacing, and real Claude/Codex harness smoke checks. Use when Reflect seems stuck, drain is not clearing pending reflections, ingest has gone quiet, statusline ERR needs proof, or a change needs tmux-backed validation.

Installation

Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.

Run Skill in Manus

Source

stevengonsalvez

stevengonsalvez/ainb-reflect-memory

View GitHub Repository View Creator Repositories

Download

Run Skill in Manus

Related occupationsSOC

Based on SOC occupation classification

Software DevelopersComputer and Mathematical Occupations·SOC 15-1252

File Explorer

2 files

SKILL.md

readonly

name	reflect-verify
description	Verify Reflect end-to-end health, especially drain, ingest/consolidate maintenance, launchd timers, statusline error surfacing, and real Claude/Codex harness smoke checks. Use when Reflect seems stuck, drain is not clearing pending reflections, ingest has gone quiet, statusline ERR needs proof, or a change needs tmux-backed validation.

Reflect Verify

Run Reflect validation in layers: static checks, isolated queue/drain smoke, live read-only health, then optional real Claude tmux smoke. Do not run full ingest or consolidate as part of verification unless the user explicitly asks; the watchdog only surfaces due/malfunction states.

Workflow

Check source and install drift.
- Compare repo hook files against installed harness copies and Claude plugin cache.
- Confirm launchd labels for com.reflect.drain, com.reflect.idle, and com.reflect.maintenance when running on macOS.
- Never kill tmux server or broad process names; only use exact session names created by this verification.

Run static gates.

.agents/skills/reflect-verify/scripts/verify-reflect.sh --static

Run isolated drain smoke.
```
.agents/skills/reflect-verify/scripts/verify-reflect.sh --isolated
```
This uses a temp REFLECT_STATE_DIR and a stub Claude binary. It must prove timeout/no-output entries are quarantined and removed from the queue without touching live Reflect state.
Run live read-only health.
```
.agents/skills/reflect-verify/scripts/verify-reflect.sh --live-readonly
```
Report queue depth, drain lock status, latest drain activity, unacked errors, ingest log age, and loaded launchd labels. Do not mutate queue, errors, or KB.

Run maintenance watchdog smoke when statusline surfacing is in scope.

REFLECT_STATE_DIR="$(mktemp -d)" \
  GLOBAL_LEARNINGS_PATH="$(mktemp -d)" \
  REFLECT_WATCH_FORCE_JSON_ERRORS=1 \
  REFLECT_WATCH_SKIP_LAUNCHD=1 \
  plugin/hooks/reflect-maintenance-watch.sh

Verify errors.json receives ingest_never_ran or drain malfunction entries.

Run real Claude tmux smoke only when needed.
```
.agents/skills/reflect-verify/scripts/verify-reflect.sh \
  --real-claude --workdir ~/d/git/ai-coder-rules
```
The script creates a tmux session named reflect-verify-<timestamp>, runs a tiny claude -p probe, captures a log, waits with a timeout, and kills only that exact session if cleanup is needed.

Done Criteria

Static checks pass.
Isolated drain smoke proves bad entries cannot head-of-line block the queue.
Live read-only check explains current Reflect state with exact queue depth, launchd labels, and latest drain/error timestamps.
If real Claude smoke is run, a tmux log proves claude -p produced a result or captures the exact failure mode.
Final report lists commands and pass/fail evidence, not just conclusions.

Safety Rules

Do not run full /reflect:ingest or /reflect:consolidate inside automated smoke checks.
Do not delete queue files. Use isolated temp state for destructive proofs.
Do not use tmux kill-server, pkill tmux, wildcard tmux kills, or process name kills.
Prefer read-only live checks unless the user explicitly asks to repair live state.

More from this repository

same repository

reflect-corpus

stevengonsalvez/ainb-reflect-memory

Build a long-lived, filtered SLICE of the global learnings KB and answer a question-set over it ("ask the auth subsystem", "ask the migration log"). Unlike /reflect:recall (a fresh hybrid query each time), a corpus is a saved filter snapshotted to disk — prime it once, ask many questions, reprime when the KB drifts. Use when the user wants a durable Q&A session against one filtered area of their own code history rather than a one-shot search.

2026-07-171

reflect-cost

stevengonsalvez/ainb-reflect-memory

Report reflect drain spend over a time window — tokens split by cached (cache_read), uncached writes (cache_creation), and io (input+output), with a $ estimate, grouped by day / outcome / model / transcript. Reads the drainer's cost log and surfaces outlier runs and cache-reuse health (the 41.5M-token failure mode = low cache reuse + high cache writes). Also reports the recall followup rate (the recall-quality self-monitor: how often a session searched again within 30s and got disjoint results). Use to answer "what is reflection costing me" for the last day / week, or "is recall satisfying".

2026-07-171

reflect-export

stevengonsalvez/ainb-reflect-memory

Export or import the reflect knowledge base as a deterministic, git-friendly tarball — the learnings markdown corpus plus the reflect.db rows in one snapshot. Use to back up the KB, move it to another machine, or restore it: `export` writes the tarball, `import` restores it (with an optional GraphRAG reindex). Answers "snapshot my reflect KB", "move my learnings to another machine", or "restore my knowledge base from a backup".

2026-07-171

reflect-ingest

stevengonsalvez/ainb-reflect-memory

The global knowledge indexer. Harvests ALL memory sources across all tools (Claude, Codex, Copilot, Gemini) and all project types into the unified GraphRAG + QMD knowledge base. Archives originals, generates entity sidecars, and dual-indexes for future retrieval. This is THE command that makes the knowledge base comprehensive.

2026-07-171

reflect

stevengonsalvez/ainb-reflect-memory

Full conversation scan for self-improvement. Detects behavioral corrections and knowledge signals, classifies them, proposes agent updates and knowledge notes with entity sidecars for GraphRAG indexing. Correct once, never again.

2026-07-171

reflect-recall

stevengonsalvez/ainb-reflect-memory

Retrieve relevant prior learnings from the global knowledge base. Hybrid vector + graph search over 170+ indexed learnings, reranked by confidence, recency, and tag overlap. Use when starting work, debugging a recurring problem, or before implementing a feature that may have prior art.

2026-07-161

name	reflect-verify
description	Verify Reflect end-to-end health, especially drain, ingest/consolidate maintenance, launchd timers, statusline error surfacing, and real Claude/Codex harness smoke checks. Use when Reflect seems stuck, drain is not clearing pending reflections, ingest has gone quiet, statusline ERR needs proof, or a change needs tmux-backed validation.

Reflect Verify

Workflow

Check source and install drift.
- Compare repo hook files against installed harness copies and Claude plugin cache.
- Confirm launchd labels for com.reflect.drain, com.reflect.idle, and com.reflect.maintenance when running on macOS.
- Never kill tmux server or broad process names; only use exact session names created by this verification.

Run static gates.

.agents/skills/reflect-verify/scripts/verify-reflect.sh --static

Run isolated drain smoke.
```
.agents/skills/reflect-verify/scripts/verify-reflect.sh --isolated
```
This uses a temp REFLECT_STATE_DIR and a stub Claude binary. It must prove timeout/no-output entries are quarantined and removed from the queue without touching live Reflect state.
Run live read-only health.
```
.agents/skills/reflect-verify/scripts/verify-reflect.sh --live-readonly
```
Report queue depth, drain lock status, latest drain activity, unacked errors, ingest log age, and loaded launchd labels. Do not mutate queue, errors, or KB.

Run maintenance watchdog smoke when statusline surfacing is in scope.

REFLECT_STATE_DIR="$(mktemp -d)" \
  GLOBAL_LEARNINGS_PATH="$(mktemp -d)" \
  REFLECT_WATCH_FORCE_JSON_ERRORS=1 \
  REFLECT_WATCH_SKIP_LAUNCHD=1 \
  plugin/hooks/reflect-maintenance-watch.sh

Verify errors.json receives ingest_never_ran or drain malfunction entries.

Run real Claude tmux smoke only when needed.
```
.agents/skills/reflect-verify/scripts/verify-reflect.sh \
  --real-claude --workdir ~/d/git/ai-coder-rules
```
The script creates a tmux session named reflect-verify-<timestamp>, runs a tiny claude -p probe, captures a log, waits with a timeout, and kills only that exact session if cleanup is needed.

Done Criteria

Static checks pass.
Isolated drain smoke proves bad entries cannot head-of-line block the queue.
Live read-only check explains current Reflect state with exact queue depth, launchd labels, and latest drain/error timestamps.
If real Claude smoke is run, a tmux log proves claude -p produced a result or captures the exact failure mode.
Final report lists commands and pass/fail evidence, not just conclusions.

Safety Rules

Do not run full /reflect:ingest or /reflect:consolidate inside automated smoke checks.
Do not delete queue files. Use isolated temp state for destructive proofs.
Do not use tmux kill-server, pkill tmux, wildcard tmux kills, or process name kills.
Prefer read-only live checks unless the user explicitly asks to repair live state.