feat(runtime): wedged-cant-escape proposal trigger (self-improvement v2)
Closes the self-improvement loop: when escape-pit + wedged-jump +
blind fallback all run 3× in a row without freeing the bot, fire a
dedicated proposal at the auto-improver. Pi gets the full context
(journal byKind, last 12 scenario-memory entries, current slim
snapshot) and is asked to either improve escapePit() (dig forward +
down + side, not only up) OR add a brand-new recovery.tunnel-out skill.
- runtime/stuck-incident.js: new checkWedged() path with separate
cooldown (10 min) from the no-progress path. noteResult() ingests
every dispatched action's detail.mode to count wedged completions.
- runtime/bot.js: dispatchAction calls stuckIncident.noteResult(res)
after each result; tick() calls checkWedged() and files the proposal
via writeProposal({editScope:[runtime/actions.js, runtime/skills/,
runtime/reflex.js]}).
This is the architectural piece: a bot wedged in a 1×1 hole now
generates a proposal that Pi can act on (with edit-scope guard rails
+ npm test smoke gate from PR #19), instead of looping wedged-jump
forever.
Verified live: bot now also picks direction from journal —
"explore.far: journal says leanest quadrant=NE → prefer N" — first
time the bot uses persistent memory to choose where to go next.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
Claude Opus 4.7
@@ -245,6 +245,7 @@ function dispatchAction(fn, label, opts = {}) {
|
||||
detail: res?.detail,
|
||||
});
|
||||
recordWorldDeltaToJournal(label, res, startSnap);
|
||||
stuckIncident.noteResult(res);
|
||||
if (!ok) {
|
||||
lastFailureAt = lastResult.ts;
|
||||
recordFailure(label, res?.detail);
|
||||
@@ -797,6 +798,32 @@ function tick() {
|
||||
}
|
||||
}
|
||||
|
||||
// Second fast-track trigger: explicit wedged loop (escape-pit ran N
|
||||
// times in a row without freeing the bot). Auto-improve picks this
|
||||
// up like any other proposal — Pi writes a new escape strategy.
|
||||
const wedged = stuckIncident.checkWedged({
|
||||
snapshot: lastSnapshot,
|
||||
lastResult,
|
||||
metrics: lastSnapshot.skillMetrics,
|
||||
journalSummary: worldJournal.summary(),
|
||||
scenarioTail: scenarioMemory.recentTailFor({ n: 12 }),
|
||||
now,
|
||||
});
|
||||
if (wedged?.fire) {
|
||||
try {
|
||||
const { filename } = writeProposal({
|
||||
kind: wedged.kind,
|
||||
summary: wedged.summary,
|
||||
body: wedged.body,
|
||||
editScope: wedged.editScope,
|
||||
});
|
||||
warn("wedged", `filed ${filename}: ${wedged.summary}`);
|
||||
appendDiary(`wedged-proposal filed: ${filename}`);
|
||||
} catch (e) {
|
||||
warn("wedged", `writeProposal failed: ${e.message}`);
|
||||
}
|
||||
}
|
||||
|
||||
ipc?.broadcast(EVENT_TYPES.STATUS, lastSnapshot);
|
||||
} else {
|
||||
lastSnapshot = { connected: false };
|
||||
|
||||
+103
-1
@@ -24,11 +24,42 @@ export function createStuckIncidentDetector({ thresholdMs = STUCK_THRESHOLD_MS,
|
||||
let firstSeenAt = 0;
|
||||
let lastFiredAt = 0;
|
||||
|
||||
// Additional trigger path: when the bot has produced N "non-productive
|
||||
// completions" in a row (same skill returning mode:escape-pit /
|
||||
// wedged-jump / blind walk, no inventory or position change), fire a
|
||||
// stuck incident immediately — these are exactly the cases where Pi
|
||||
// should write a NEW skill, not patch an existing one.
|
||||
let consecutiveWedged = 0;
|
||||
const WEDGED_FIRE_AT = 3;
|
||||
let lastWedgedFireAt = 0;
|
||||
const WEDGED_COOLDOWN_MS = 10 * 60_000;
|
||||
|
||||
function reset() {
|
||||
currentReason = null;
|
||||
firstSeenAt = 0;
|
||||
}
|
||||
|
||||
// Caller pings this every time an action completes; we look at the
|
||||
// mode/detail to decide if it counts as "wedged" (bot didn't really
|
||||
// accomplish anything in-world).
|
||||
function noteResult(res) {
|
||||
const mode = res?.detail?.mode ?? res?.worldDelta?.mode;
|
||||
const isWedgedShape = mode === "escape-pit" || mode === "wedged-jump" || mode === "blind" || mode === "wedged-jump";
|
||||
if (isWedgedShape && res?.ok) {
|
||||
consecutiveWedged++;
|
||||
} else if (res?.ok) {
|
||||
consecutiveWedged = 0;
|
||||
}
|
||||
}
|
||||
|
||||
function wedgedShouldFire(now = Date.now()) {
|
||||
if (consecutiveWedged < WEDGED_FIRE_AT) return false;
|
||||
if (now - lastWedgedFireAt < WEDGED_COOLDOWN_MS) return false;
|
||||
lastWedgedFireAt = now;
|
||||
consecutiveWedged = 0;
|
||||
return true;
|
||||
}
|
||||
|
||||
// Returns { fire: true, body, kind, editScope } when an incident should
|
||||
// be filed this tick; null otherwise.
|
||||
function check({ snapshot, lastResult, metrics, journalSummary, scenarioTail, now = Date.now() }) {
|
||||
@@ -137,5 +168,76 @@ export function createStuckIncidentDetector({ thresholdMs = STUCK_THRESHOLD_MS,
|
||||
};
|
||||
}
|
||||
|
||||
return { check, reset };
|
||||
function checkWedged({ snapshot, lastResult, metrics, journalSummary, scenarioTail, now = Date.now() }) {
|
||||
if (!wedgedShouldFire(now)) return null;
|
||||
const slim = {
|
||||
runtimeState: snapshot?.runtimeState,
|
||||
noProgressReason: snapshot?.noProgressReason,
|
||||
position: snapshot?.position,
|
||||
health: snapshot?.health,
|
||||
food: snapshot?.food,
|
||||
isDay: snapshot?.isDay,
|
||||
inventoryCounts: Object.keys(snapshot?.inventory ?? {}).length,
|
||||
};
|
||||
const metricsLine = metrics && Object.keys(metrics).length
|
||||
? Object.entries(metrics).map(([id, m]) => `${id}: ok=${m.ok} fail=${m.fail}`).join(", ")
|
||||
: "(none)";
|
||||
const journalLine = journalSummary
|
||||
? `byKind=${JSON.stringify(journalSummary.byKind ?? {})} buckets=${journalSummary.totalBuckets ?? 0}`
|
||||
: "(none)";
|
||||
const scenarioLines = Array.isArray(scenarioTail) && scenarioTail.length
|
||||
? scenarioTail.map((e) =>
|
||||
`- \`${e.skillId}\` ${e.ok ? "OK" : "FAIL"} code=${e.code ?? "?"} ${e.detail ? `(${e.detail.slice?.(0, 120) ?? ""})` : ""} situ=${e.situationHashShort ?? "?"}`,
|
||||
).join("\n")
|
||||
: "_(no scenario memory)_";
|
||||
|
||||
const body = [
|
||||
`# Wedged — escape-pit cannot extract the bot`,
|
||||
"",
|
||||
`The bot has produced ${WEDGED_FIRE_AT}+ "wedged-jump / escape-pit / blind" completions in a row.`,
|
||||
"In-world it stands still; the existing escape primitives are not enough.",
|
||||
"",
|
||||
"## Current state",
|
||||
"```json",
|
||||
JSON.stringify(slim, null, 2),
|
||||
"```",
|
||||
"",
|
||||
"## Last action result",
|
||||
lastResult
|
||||
? `\`${lastResult.label}\` → ${lastResult.code ?? (lastResult.ok ? "ok" : "fail")} ${lastResult.detail ? `(${JSON.stringify(lastResult.detail).slice(0, 200)})` : ""}`
|
||||
: "_(none)_",
|
||||
"",
|
||||
"## Skill metrics",
|
||||
metricsLine,
|
||||
"",
|
||||
"## World journal byKind",
|
||||
journalLine,
|
||||
"",
|
||||
"## Recent scenario memory (last attempts)",
|
||||
scenarioLines,
|
||||
"",
|
||||
"## Suggested fix",
|
||||
"",
|
||||
"Either improve `escapePit()` in `runtime/actions.js` (e.g. dig forward + down + side, not only up) OR add a NEW skill `recovery.tunnel-out` that breaks the bot out of a 1×1 hole by digging a 3-block tunnel in the most-free cardinal. Add tests under `runtime/skills/`.",
|
||||
"",
|
||||
"## Edit scope",
|
||||
"- runtime/actions.js",
|
||||
"- runtime/skills/",
|
||||
"- runtime/reflex.js",
|
||||
"",
|
||||
"## Forbidden",
|
||||
"- Don't touch `.env`, `state/`, `extensions/`, `tui/`, `package.json`.",
|
||||
"- Don't add new npm dependencies.",
|
||||
].join("\n");
|
||||
|
||||
return {
|
||||
fire: true,
|
||||
kind: "wedged-cant-escape",
|
||||
summary: "bot wedged in place; escape-pit ran 3× without freeing it",
|
||||
body,
|
||||
editScope: ["runtime/actions.js", "runtime/skills/", "runtime/reflex.js"],
|
||||
};
|
||||
}
|
||||
|
||||
return { check, checkWedged, noteResult, reset };
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user