pi-goal-list-loop-audit 0.38.13 → 0.38.15

package/CHANGELOG.md

@@ -1,5 +1,15 @@
 # Changelog
 
+## 0.38.15 — pause anti-confabulation (2026-09-04)
+
+### Changed
+  `pause_goal` refuses pauses whose blocker claims a GLLA tool is missing (dispatch proves the batch landed) and tells the model to call it; quoting pi's `Tool X not found` is accepted as genuine outage. Runbook + new-tab forensics in `audit/PAUSE-ANTI-CONFABULATION-2026-09-04.md`.
+
+## 0.38.14 — human end-of-objective briefing (2026-09-04)
+
+### Changed
+  Every `✓ done` chat notify is now a human briefing: outcome first in its own words, only informing labels after it (`none` / `not recorded` / empty filler dropped; `none — <content>` keeps the content). Word-boundary cuts, auditor/no-audit trailer, archive keeps the full six-label record. Runbook in `audit/HUMAN-BRIEF-2026-09-04.md`.
+
 ## 0.38.13 — good completion summary (2026-09-04)
 
 ### Changed

package/docs/INDEX.md

@@ -18,7 +18,7 @@ For shipped docs, the relevant entry points are:
   failback; v0.35.9 hardened cross-version npm tarball checks; v0.35.10
   handles multi-entry npm dry-run reports; v0.35.11 accepts both npm report
   shapes; v0.35.12 supports npm 12's keyed pack reports; v0.35.13 fixes stale-API recovery loops.
-  v0.35.14–v0.38.13 continue through the supervisor freeze (`/glla pause`),
+  v0.35.14–v0.38.15 continue through the supervisor freeze (`/glla pause`),
   load hold, auditor picker parity, Windows launch fix, zombie-watchdog
   subagent carve-out, due-wait backstop, the `/glla agents` visibility panel,
   durable state-root selection, blank-until-resume auditor context, frozen

package/extensions/completion-summary.ts

@@ -116,6 +116,49 @@ export function compactCompletionSummary(text: string | undefined, maxValueLengt
   return parts.join(" · ");
 }
 
+/** A recap value informs the human only when it is not an empty / none /
+ * not-recorded placeholder. `None — <real content>` (a common agent habit)
+ * keeps just the content. Returns null for filler. */
+export function briefValueContent(value: string): string | null {
+  const clean = value.replace(/\s+/g, " ").trim();
+  if (!clean) return null;
+  // System placeholder vocabulary: the `— explanation` is still a placeholder.
+  if (/^not recorded\b/i.test(clean)) return null;
+  // Agent habit: `none — <real content>` hides information behind filler.
+  const prefixed = /^none\s*[—–:\-]\s*(.+)$/i.exec(clean);
+  const body = (prefixed?.[1] ?? clean).trim();
+  if (/^(none|n\/a|nil|nothing)(\s+for\s+this\s+\w+)?\.?$/i.test(body)) return null;
+  return body;
+}
+
+export interface HumanCompletionBrief {
+  outcome: string;
+  details: string[];
+}
+
+/** The human briefing: outcome first in its own words, then only the
+ * labels that carry real content (filler like `Unresolved: none` or
+ * `Changed: not recorded` is dropped, never shown). The durable archive
+ * keeps the full six-label record; this is the end-of-objective voice
+ * that informs the user at a glance. */
+export function humanCompletionBrief(
+  text: string | undefined,
+  outcomeBudget = 140,
+  valueBudget = 120,
+): HumanCompletionBrief {
+  const lines = completionSummaryLines(text, Math.max(outcomeBudget, valueBudget));
+  const rawOutcome = (lines[0] ?? "").replace(/^Outcome:\s*/, "");
+  const outcome = clipSummaryValue(briefValueContent(rawOutcome) ?? "done", outcomeBudget);
+  const details: string[] = [];
+  for (const line of lines.slice(1)) {
+    const separator = line.indexOf(":");
+    if (separator < 0) continue;
+    const content = briefValueContent(line.slice(separator + 1));
+    if (content) details.push(`${line.slice(0, separator)}: ${clipSummaryValue(content, valueBudget)}`);
+  }
+  return { outcome, details };
+}
+
 /** Multi-line projection: one `Label: value` line per label with generous
  * word-bounded values. This is the user-facing `✓ done` block — six short
  * facts that stay scannable in chat. The single-line projection remains
@@ -236,6 +279,15 @@ export function compactTerminalCompletionSummary(
   return compactCompletionSummary(resolveCompletionSummary(facts, candidate).summary, maxValueLength);
 }
 
+/** Brief twin of compactTerminalCompletionSummary for the `✓ done` chat
+ * notifies: same resolved facts, outcome + informing labels only. */
+export function terminalHumanBrief(
+  facts: CompletionSummaryFacts,
+  candidate = facts.goal.completionSummary,
+): HumanCompletionBrief {
+  return humanCompletionBrief(resolveCompletionSummary(facts, candidate).summary);
+}
+
 /** Multi-line twin of compactTerminalCompletionSummary for the `✓ done`
  * chat notifies: same resolved facts, one `Label: value` line each. */
 export function terminalCompletionSummaryLines(

package/extensions/goal-loop-core.ts

@@ -3056,6 +3056,28 @@ export function missingGllaTools(activeNames: readonly string[]): readonly GllaT
   return GLLA_TOOL_NAMES.filter((n) => !active.has(n));
 }
 
+/** pi's own wording when a model calls a tool pi cannot dispatch. A pause
+ * reason quoting this is evidence of a genuine tool outage — the pause
+ * must be accepted, never refused as confabulation. */
+export const PI_TOOL_NOT_FOUND_QUOTE = /tool\s+\S+\s+not found/i;
+
+/** A pause reason that claims a GLLA tool does not exist in this session
+ * (`no complete_goal tool`, `complete_goal is missing`, …). Ordinary
+ * mentions of a tool (results, instructions, summaries) never match —
+ * absence language is required. */
+export function claimedMissingGllaTool(reason: string): GllaToolName | null {
+  const text = (reason ?? "").replace(/\s+/g, " ");
+  if (!text) return null;
+  for (const name of GLLA_TOOL_NAMES) {
+    const absence = new RegExp(
+      `\\bno\\s+${name}\\s+tool\\b|\\b${name}\\b[^.]{0,120}?\\b(missing|not found|not available|unavailable|doesn'?t exist|isn'?t available|is not available|no longer (?:available|exists|present))\\b`,
+      "i",
+    );
+    if (absence.test(text)) return name;
+  }
+  return null;
+}
+
 // -----------------------------------------------------------------
 // v0.25.0 — eager-continuation contract helpers
 // -----------------------------------------------------------------

package/extensions/loops/goal-auditor-hooks.ts

@@ -240,7 +240,7 @@ import {
   pushCapped as pushRepetitionCapped,
 } from "../goal-loop-repetition.js";
 import { buildStatusText, buildWidgetLines, type AuditDisplayProgress } from "../goal-loop-display.js";
-import { compactCompletionSummary, compactTerminalCompletionSummary, isGenericCompletionSummary, missingCompletionSummaryLabels, terminalCompletionSummaryLines } from "../completion-summary.js";
+import { compactCompletionSummary, compactTerminalCompletionSummary, isGenericCompletionSummary, missingCompletionSummaryLabels, terminalHumanBrief } from "../completion-summary.js";
 import {
   defaultAgentDir,
   resolveEffectiveSubagentModel,
@@ -1416,7 +1416,7 @@ async function retryStoredCompletionAudit(origin: CompletionAuditOrigin = "provi
       archivePath: path.relative(liveCtx.cwd, archivedGoalPath(liveCtx.cwd, state.goal.id)) || archivedGoalPath(liveCtx.cwd, state.goal.id),
     }, state.goal.completionSummary);
     // Computed pre-archive: archiveCurrentGoal clears state.goal.
-    const recapLines = terminalCompletionSummaryLines({
+    const brief = terminalHumanBrief({
       goal: state.goal,
       status: "complete",
       stopReason: terminalReason,
@@ -1438,10 +1438,10 @@ async function retryStoredCompletionAudit(origin: CompletionAuditOrigin = "provi
       appendLedger(liveCtx.cwd, "goal_archive_failed_after_approval", { goalId, attemptId: claim.attemptId, origin });
       return;
     }
-    // v0.38.13: the chat notify carries the six-label block (one fact per
-    // line, word-bounded) instead of the single-line mash — the external
-    // notify keeps the compact single line (pager/sound safe).
-    liveCtx.ui.notify(`✓ done — auditor ${result.model} approved${approvalVia}.\n${recapLines.join("\n")}`, "info");
+    // v0.38.14: the chat notify is the human briefing — outcome first,
+    // filler labels dropped — while the external notify keeps the compact
+    // single line (pager/sound safe).
+    liveCtx.ui.notify(`✓ done — ${brief.outcome}\n${[...brief.details, `— auditor ${result.model} approved${approvalVia}.`].join("\n")}`, "info");
     notifyExternal(liveCtx, `Goal complete (auditor approved, ${origin}): ${recap}`);
     return;
   }

package/extensions/loops/goal-tools.ts

@@ -114,6 +114,8 @@ import {
   readQueueFromDisk,
   deleteQueueItemFileResult,
   missingGllaTools,
+  claimedMissingGllaTool,
+  PI_TOOL_NOT_FOUND_QUOTE,
   runPersistStep,
   isPersistenceDegraded,
   lastPersistenceFailure,
@@ -250,7 +252,7 @@ import {
   pushCapped as pushRepetitionCapped,
 } from "../goal-loop-repetition.js";
 import { buildStatusText, buildWidgetLines, type AuditDisplayProgress } from "../goal-loop-display.js";
-import { compactCompletionSummary, compactTerminalCompletionSummary, resolveCompletionSummary, terminalCompletionSummaryLines } from "../completion-summary.js";
+import { compactCompletionSummary, compactTerminalCompletionSummary, resolveCompletionSummary, terminalHumanBrief } from "../completion-summary.js";
 import {
   defaultAgentDir,
   resolveEffectiveSubagentModel,
@@ -1216,15 +1218,16 @@ function registerAgentTools(pi: any): void {
               details: {},
             };
           }
-          const recapLines = terminalCompletionSummaryLines({
+          const brief = terminalHumanBrief({
             goal: terminalGoal,
             status: "complete",
             stopReason: terminalReason,
             archivePath: path.relative(ctx.cwd, archivedGoalPath(ctx.cwd, terminalGoal.id)) || archivedGoalPath(ctx.cwd, terminalGoal.id),
           });
-          ctx.ui.notify(`✓ done without audit (user choice):\n${recapLines.join("\n")}`, "info");
+          const briefBlock = [...brief.details, `— completed without audit (your choice).`].join("\n");
+          ctx.ui.notify(`✓ done — ${brief.outcome}\n${briefBlock}`, "info");
           notifyExternal(ctx, `Goal complete without audit (user choice): ${recap}`);
-          return { content: [{ type: "text", text: `Goal marked complete without audit (user choice).\n\n${recapLines.join("\n")}` }], details: {} };
+          return { content: [{ type: "text", text: `Goal marked complete without audit (user choice).\n\n${briefBlock}` }], details: {} };
         }
         scheduleContinuation(ctx, true);
         return {
@@ -1250,7 +1253,7 @@ function registerAgentTools(pi: any): void {
           archivePath: path.relative(ctx.cwd, archivedGoalPath(ctx.cwd, state.goal.id)) || archivedGoalPath(ctx.cwd, state.goal.id),
         }, state.goal.completionSummary);
         // Computed pre-archive: archiveCurrentGoal clears state.goal.
-        const recapLines = terminalCompletionSummaryLines({
+        const brief = terminalHumanBrief({
           goal: state.goal,
           status: "complete",
           stopReason: terminalReason,
@@ -1271,7 +1274,7 @@ function registerAgentTools(pi: any): void {
           appendLedger(ctx.cwd, "goal_archive_failed_after_approval", { goalId: state.goal?.id, origin: "manual-verify", model: result.model });
           return { content: [{ type: "text", text: "The auditor approved, but the terminal archive could not be persisted. The goal is paused; fix persistence, resume, and retry complete_goal." }], details: {} };
         }
-        ctx.ui.notify(`✓ done — auditor ${result.model} approved.\n${recapLines.join("\n")}`, "info");
+        ctx.ui.notify(`✓ done — ${brief.outcome}\n${[...brief.details, `— auditor ${result.model} approved.`].join("\n")}`, "info");
         notifyExternal(ctx, `Goal complete (auditor approved): ${recap}`);
         return { content: [{ type: "text", text: `Goal approved by auditor ${result.model}.` }], details: {} };
       }
@@ -1926,6 +1929,20 @@ function registerAgentTools(pi: any): void {
       if (state.goal.status !== "active") {
         return { content: [{ type: "text", text: `Goal is already ${state.goal.status}; pause request ignored.` }], details: {} };
       }
+      // v0.38.15: anti-confabulation — a pause whose blocker is "this
+      // session has no <glla tool>" is refused when the tool path provably
+      // works: this very pause_goal call dispatched through the same
+      // registration batch (complete_goal registers first in that batch),
+      // so the tool IS callable — the model is misreading its tool list
+      // (field: new-tab 2026-09-04, 5×-compacted session, zero tool errors
+      // in transcript or ledger). The model is told to call it now; a
+      // quoted pi `Tool X not found` error is genuine-outage evidence and
+      // the pause is accepted.
+      const missingClaim = claimedMissingGllaTool(p.reason ?? "");
+      if (missingClaim && !PI_TOOL_NOT_FOUND_QUOTE.test(p.reason ?? "")) {
+        appendLedger(ctx.cwd, "pause_refused_tool_present", { goalId: state.goal.id, tool: missingClaim });
+        return { content: [{ type: "text", text: `Not paused: \`${missingClaim}\` is registered in this session — this pause_goal call just dispatched through the same registration batch, so the tool path works. If \`${missingClaim}\` is missing from your visible tool list, that is a client-side gap: call \`${missingClaim}\` now${missingClaim === "complete_goal" ? " with your six-label recap and verification summary" : ""}. If pi itself answers with a \`Tool ${missingClaim} not found\` error, call pause_goal again quoting that exact error and the pause will be accepted.` }], details: {} };
+      }
       const pauseCopy = providerErrorPresentation(p.reason, "recovery");
       const safePauseReason = pauseCopy.sensitive ? pauseCopy.display : p.reason;
       const safePauseAction = p.suggestedAction ? sanitizeProviderDisplayText(p.suggestedAction) : p.suggestedAction;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "pi-goal-list-loop-audit",
-  "version": "0.38.13",
+  "version": "0.38.15",
   "description": "Mission control for autonomous pi: interview-drafted goals, an audited task queue, and forever-loops (metric, spec, project-audit) that run for hours. A detached extension-less auditor process re-verifies every completion with raw evidence without holding the main pi turn; confirmed drafts, decision pauses and consent gates keep you in charge.",
   "license": "AGPL-3.0-only",
   "author": "dracon",