@linimin/pi-letscook 0.1.70 → 0.1.71

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/CHANGELOG.md CHANGED
@@ -1,5 +1,12 @@
1
1
  # Changelog
2
2
 
3
+ ## 0.1.71
4
+
5
+ ### Changed
6
+
7
+ - clarified the packaged `completion-auditor` output contract so `Stale or conflicting canonical state` must begin with `yes` or `no`, matching the canonical transcription gate
8
+ - added rubric-contract coverage to keep the stricter auditor yes/no guidance from drifting and to reduce avoidable transcription warnings during audit
9
+
3
10
  ## 0.1.70
4
11
 
5
12
  ### Changed
@@ -52,9 +52,11 @@ Answer only:
52
52
  - `Tracked and unignored worktree is clean: yes/no`
53
53
  - `Worktree blockers: ...`
54
54
  - `Next mandatory slice: ...`
55
- - `Stale or conflicting canonical state: ...`
55
+ - `Stale or conflicting canonical state: yes/no - ...`
56
56
  - `Plan truthfully captures remaining slice backlog: yes/no - ...`
57
57
 
58
+ For every yes/no audit field, start the value with exactly `yes` or `no`. Do not substitute `none`, `clear`, `fresh`, `unknown`, or other synonyms. For example: `Stale or conflicting canonical state: no - canonical state remains aligned with the active slice and backlog.`
59
+
58
60
  If the tracked and unignored worktree is dirty after the latest committed slice, report that as a blocker to next-slice progression, do not recommend a new next slice, and point the workflow back to reconciliation of the latest slice.
59
61
 
60
62
  If no remaining gap is evident, say so plainly instead of inventing one.
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@linimin/pi-letscook",
3
- "version": "0.1.70",
3
+ "version": "0.1.71",
4
4
  "description": "Pi package for long-running completion workflows with canonical .agent state, role-based subagents, continuity, and verification helpers.",
5
5
  "license": "MIT",
6
6
  "private": false,
@@ -90,6 +90,8 @@ assertIncludes('extensions/completion/role-reporting.js', 'Reviewer output canno
90
90
  assertIncludes('extensions/completion/role-reporting.js', 'Auditor output must answer \'Tracked and unignored worktree is clean\' with yes or no.');
91
91
  assertIncludes('extensions/completion/role-reporting.js', 'Auditor output must answer \'Stale or conflicting canonical state\' with yes or no.');
92
92
  assertIncludes('extensions/completion/role-reporting.js', 'Auditor output must answer \'Plan truthfully captures remaining slice backlog\' with yes or no.');
93
+ assertIncludes('agents/completion-auditor.md', '`Stale or conflicting canonical state: yes/no - ...`');
94
+ assertIncludes('agents/completion-auditor.md', 'For every yes/no audit field, start the value with exactly `yes` or `no`.');
93
95
  assertIncludes('extensions/completion/role-reporting.js', 'Stop-judge output cannot mark \'Can the project stop now: yes\' when any rubric line is fail.');
94
96
  assertIncludes('extensions/completion/role-reporting.js', 'Stop-judge output must answer \'Docs/config/runbooks match shipped behavior\' with yes or no.');
95
97
  assertIncludes('extensions/completion/role-reporting.js', 'Stop-judge output must answer \'Tracked and unignored worktree is clean\' with yes or no.');