fft-debug
Skill4 starsChanged 8 months agoMITView on GitHub
Diagnose and resolve FFT_nano runtime issues across container execution, provider credentials, Telegram/WhatsApp routing, scheduler behavior, and per-group logs/state.
In this repository23 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- add-telegramSkill
- add-telegram-swarmSkill
- agent-browserSkill
- agent-harness-engineerSkill
- autoresearch-createSkill
- autoresearch-finalizeSkill
- fft-coder-opsSkill
- fft-dashboard-opsSkill
- fft-farm-bootstrapSkill
- fft-farm-onboardingSkill
- fft-farm-opsSkill
- fft-host-ipcSkill
- fft-setupSkill
- fft-telegram-opsSkill
- loopsSkill
- rapid-researchSkill
- skill-mechanicSkill
- web-searchSkill
- customizeSkill
- debugSkill
- setupSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
14 rules. About 0 are common, 11 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈11
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 8
Use this skill for incident triage and deterministic debugging of FFT_nano host and container paths.
- Distinctive: line 12
Use when runtime behavior is failing, degraded, or inconsistent.
- Distinctive: line 13
Use when investigating provider wiring, routing, logs, or IPC state.
- Distinctive: line 14
Use when producing reproducible triage output before any fix attempt.
- Distinctive: line 18
Do not use as the default path for normal feature work.
- Distinctive: line 19
Do not use when the request is purely product behavior, not incident/debug.
- Distinctive: line 20
Do not use to justify destructive cleanup without explicit approval.
- Distinctive: line 24
Never use destructive git commands unless explicitly requested.
- Distinctive: line 26
Keep admin/coder controls in main chat only.
- Distinctive: line 32
Check startup/runtime logs:
- Distinctive: line 35
Check per-group container logs:
- Distinctive: line 61
Verify bot token present at runtime.
- Very specific: line 64
If /main <secret> fails, verify TELEGRAM_ADMIN_SECRET and restart.
- Very specific: line 72
verify explicit flags: CONTAINER_RUNTIME=host, FFT_NANO_ALLOW_HOST_RUNTIME=1
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
FFT Debug
Use this skill for incident triage and deterministic debugging of FFT_nano host and container paths.
When to use this skill
- Use when runtime behavior is failing, degraded, or inconsistent.
- Use when investigating provider wiring, routing, logs, or IPC state.
- Use when producing reproducible triage output before any fix attempt.
When not to use this skill
- Do not use as the default path for normal feature work.
- Do not use when the request is purely product behavior, not incident/debug.
- Do not use to justify destructive cleanup without explicit approval.
Guardrails
- Never use destructive git commands unless explicitly requested.
- Preserve unrelated worktree changes.
- Keep admin/coder controls in main chat only; do not bypass main-chat restrictions.
Fast Triage
- Confirm single instance lock:
ls -la data/fft_nano.lock
- Check startup/runtime logs:
tail -n 200 logs/fft_nano.logtail -n 200 logs/fft_nano.error.log
- Check per-group container logs:
ls -la groups/<group>/logstail -n 200 groups/<group>/logs/*
- Inspect registration + IPC state:
cat data/registered_groups.jsonfind data/ipc -maxdepth 3 -type f
Common Failure Modes
- Provider key missing: Pi reports no models available.
- Wrong
PI_API/PI_MODEL: model/provider not found. - Multiple bot instances: Telegram getUpdates conflict.
- Docker daemon unavailable or runtime permission issues.
Recommended Debug Commands
- Build/runtime checks:
npm run typechecknpm test
- Dry-run routing check (no live model call):
FFT_NANO_DRY_RUN=1 ./scripts/start.sh dev telegram-only
- Runtime detail logging:
LOG_LEVEL=debug ./scripts/start.sh dev telegram-only
Telegram Debug Path
- Verify bot token present at runtime.
- In bot DM:
/id,/status,/help. - Admin checks in main chat only:
/tasks,/groups,/panel,/coder,/coder-plan. - If
/main <secret>fails, verifyTELEGRAM_ADMIN_SECRETand restart.
Container Runtime Debug
- Docker runtime:
docker infodocker ps
- Host runtime (unisolated):
- verify explicit flags:
CONTAINER_RUNTIME=host,FFT_NANO_ALLOW_HOST_RUNTIME=1
- verify explicit flags:
Database/State Debug
- SQLite file location:
store/messages.db - Router state:
data/router_state.json - Group registry:
data/registered_groups.json - Per-group Pi state:
data/pi/<group>/.pi/
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

