| name | witness-observer |
| description | PMU observation pattern for agent health monitoring. Runs patrol loops, detects stalled agents, sends nudges, and escalates persistent failures to the mayor. Never modifies agent work or resolves conflicts. |
| type | skill |
| category | orchestration |
| status | stable |
| origin | tibsfox |
| modified | false |
| first_seen | "2026-03-06T00:00:00.000Z" |
| first_path | .claude/skills/witness-observer/SKILL.md |
| superseded_by | null |
Witness Observer
Per-rig observer that monitors polecat health and reports anomalies. The witness is the PMU (Performance Monitoring Unit) of the Gastown chipset -- it watches execution units for stalls, detects degraded performance, and raises alerts without interfering with computation. The witness is strictly read-only with respect to agent work. It observes and reports; it never modifies.
Activation Triggers
This skill activates when:
- The agent is assigned to monitor a rig's worker agents
- Multiple polecats are running and health monitoring is needed
- Stall detection is required for long-running work items
- The mayor needs a supervisory agent to watch active polecats
Core Capabilities
Patrol Loop
The witness runs a periodic patrol that checks all active agents in its rig for health indicators.
Patrol cycle:
SCAN EVALUATE ACT WAIT
| | | |
v v v v
list agents -> check each -> nudge/escalate -> sleep interval
(active ones) for stalls if needed (default 5 min)
Implementation:
const state = new StateManager({ stateDir: '.chipset/state/' });
const patrolInterval = 5 * 60 * 1000;
const stallThreshold = 30 * 60 * 1000;
async function patrol(): Promise<void> {
const agents = await state.listAgents({ role: 'polecat' });
active = agents.( a. === );
( agent active) {
hook = state.(agent.);
(!hook || hook. !== ) ;
lastActivity = (hook.).();
elapsed = .() - lastActivity;
(elapsed > stallThreshold) {
(agent, hook, elapsed);
}
}
}