fft-farm-ops
Skill4 starsChanged 8 months agoMITView on GitHub
Operate the farm bridge through ledger reads and IPC action requests, using main-chat-only control flows and safety guardrails.
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-debugSkill
- fft-farm-bootstrapSkill
- fft-farm-onboardingSkill
- 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
17 rules. About 0 are common, 11 are distinctive, and 6 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.
6
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 farm state awareness and operational actions routed through the FFT_nano farm bridge.
- Distinctive: line 12
Use for reading farm ledger state and alerts.
- Distinctive: line 13
Use for submitting allowlisted farm control actions through IPC.
- Distinctive: line 14
Use when main/admin chat requests operational farm actions.
- Distinctive: line 18
Do not use for onboarding/bootstrap tasks.
- Distinctive: line 19
Do not use outside main/admin chat for write/control actions.
- Distinctive: line 20
Do not use for actions outside the defined farm action allowlist.
- Distinctive: line 24
Never run destructive git commands unless explicitly requested.
- Distinctive: line 26
Keep write operations main-chat-only.
- Distinctive: line 27
Use main/admin chat only for any operational control action.
- Distinctive: line 28
Never run destructive shell commands against hosts, containers, or Home Assistant.
- Very specific: line 38
Read current.json first. Treat stale: true as degraded connectivity and avoid risky recommendations.
- Very specific: line 42
Write request files into /workspace/ipc/actions/ with unique names like:
- Very specific: line 63
Read /workspace/ipc/action_results/<requestId>.json until available.
- Very specific: line 76
If status is error, surface the exact error and propose the safest next step.
- Distinctive: line 80
Only use allowlisted actions:
- Very specific: line 130
Always include context from timeOfDay, season, and weather-related states.
≈ “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 Farm Ops
Use this skill for farm state awareness and operational actions routed through the FFT_nano farm bridge.
When to use this skill
- Use for reading farm ledger state and alerts.
- Use for submitting allowlisted farm control actions through IPC.
- Use when main/admin chat requests operational farm actions.
When not to use this skill
- Do not use for onboarding/bootstrap tasks.
- Do not use outside main/admin chat for write/control actions.
- Do not use for actions outside the defined farm action allowlist.
Guardrails
- Never run destructive git commands unless explicitly requested.
- Preserve unrelated worktree changes.
- Keep write operations main-chat-only.
- Use main/admin chat only for any operational control action.
- Never run destructive shell commands against hosts, containers, or Home Assistant.
Read Farm State
- Primary status file:
/workspace/farm-state/current.json - Alert snapshot:
/workspace/farm-state/alerts.json - Device inventory:
/workspace/farm-state/devices.json - Calendar snapshot:
/workspace/farm-state/calendar.json - Time series:
/workspace/farm-state/telemetry.ndjson
Read current.json first. Treat stale: true as degraded connectivity and avoid risky recommendations.
Submit Actions
Write request files into /workspace/ipc/actions/ with unique names like:
act_<unix_ts>_<rand>.json
Request format:
{
"type": "farm_action",
"action": "ha_call_service",
"params": {
"domain": "switch",
"service": "turn_on",
"data": { "entity_id": "switch.irrigation_north" }
},
"requestId": "act_1707830400_abc"
}Poll Action Results
Read /workspace/ipc/action_results/<requestId>.json until available.
Result format:
{
"requestId": "act_1707830400_abc",
"status": "success",
"result": {},
"executedAt": "2026-02-13T10:30:01Z"
}If status is error, surface the exact error and propose the safest next step.
Action Allowlist
Only use allowlisted actions:
ha_get_statusha_call_serviceha_set_entityha_restartha_apply_dashboardha_capture_screenshotha_dashboard_getha_dashboard_patchha_dashboard_validateha_canvas_get_specha_canvas_set_specha_canvas_patch_specfarm_state_refresh
Minimal Examples
{
"type": "farm_action",
"action": "ha_dashboard_patch",
"params": {
"dashboardFile": "/workspace/dashboard/ui-lovelace-staging.yaml",
"operations": [
{ "op": "set_theme", "theme": "storm" }
]
},
"requestId": "act_1707830400_dash"
}{
"type": "farm_action",
"action": "ha_canvas_patch_spec",
"params": {
"operations": [
{ "op": "set_title", "title": "Agent Canvas" }
]
},
"requestId": "act_1707830400_canvas"
}Alert Interpretation
- Critical alerts: prioritize immediate stabilization actions and ask for confirmation before control changes.
- Warning alerts: recommend monitored corrective actions.
- Normal state: suggest optimization and preventative maintenance.
- Always include context from
timeOfDay,season, and weather-related states.
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.

