ai-wiki-toolkit-linux-arm64 0.1.26 → 0.1.28

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/README.md CHANGED
@@ -1,7 +1,7 @@
1
1
  # ai-wiki-toolkit-linux-arm64
2
2
 
3
3
  This package contains the `aiwiki-toolkit` executable for `linux-arm64-glibc`.
4
- It is published as the platform-specific binary package for `ai-wiki-toolkit` `0.1.26`.
4
+ It is published as the platform-specific binary package for `ai-wiki-toolkit` `0.1.28`.
5
5
  Most users should install `ai-wiki-toolkit` instead of using this package directly.
6
6
 
7
7
  ---
@@ -110,6 +110,7 @@ The command emits a transient AI Wiki Context Packet with:
110
110
 
111
111
  - a coarse task type and risk tags
112
112
  - an effort level so simple operational tasks can stay lightweight
113
+ - generated success criteria and verification checks so non-trivial tasks start with a clear finish line
113
114
  - index cards with short descriptions and reference links for relevant memory
114
115
  - `must_load` docs to consult first when direct context is required
115
116
  - source-cited `must_follow` rules extracted from authoritative user-owned docs
@@ -420,10 +421,17 @@ To inspect memory quality from local reuse and task-check evidence:
420
421
  ```bash
421
422
  aiwiki-toolkit diagnose memory
422
423
  aiwiki-toolkit diagnose memory --since 14d --handle your-handle
424
+ aiwiki-toolkit diagnose memory --focus trial-error
423
425
  ```
424
426
 
425
427
  This writes regenerated local reports under `ai-wiki/_toolkit/diagnostics/` and prints the report to stdout. The report highlights high-ROI memory, noisy memory, stale or missing docs, conflict notes, missed-memory signals, and coverage gaps such as document reuse events that were never paired with a task-level reuse check. It does not edit user-owned AI wiki docs.
426
428
 
429
+ Use `--focus trial-error` to generate a focused trial/error reduction report from existing
430
+ AI wiki evidence. It summarizes material effects such as `avoided_retry`,
431
+ `blocked_wrong_path`, `changed_plan`, and `faster_resolution`, separates missed or repeated issue
432
+ signals from unproven wiki use, and lists replay candidates that still need source incident
433
+ artifacts before becoming formal impact-eval families.
434
+
427
435
  To turn diagnostics and handle-local drafts into a human-reviewable consolidation queue:
428
436
 
429
437
  ```bash
@@ -433,6 +441,28 @@ aiwiki-toolkit consolidate queue --since 14d --handle your-handle
433
441
 
434
442
  This writes regenerated local reports under `ai-wiki/_toolkit/consolidation/` and prints the queue to stdout. The queue suggests one action per draft cluster: keep, refine, promotion candidate, conflict, or supersession. It does not edit user-owned AI wiki docs or create shared conventions, review patterns, problems, features, or decisions; those still require human confirmation.
435
443
 
444
+ To summarize first-attempt product impact from a captured eval run:
445
+
446
+ ```bash
447
+ aiwiki-toolkit eval impact report --run-dir /path/to/eval-run
448
+ aiwiki-toolkit eval impact report --run-dir /path/to/eval-run --format json
449
+ aiwiki-toolkit eval impact summarize --run-dir /path/to/eval-run --run-dir /path/to/another-run
450
+ aiwiki-toolkit eval impact summarize --runs-file evals/impact/runs.json
451
+ ```
452
+
453
+ This reads an existing run directory with `metadata.json`, result captures, optional
454
+ `score.json` files, and optional `confounds.json`. It compares the run's primary variants,
455
+ normally `no_aiwiki_workflow` versus `aiwiki_ambient_memory_workflow`, using first-attempt
456
+ metrics only: `first_pass` captures count toward the signal, while `final` repair captures
457
+ stay diagnostic. The command reports first-attempt success rate, average score, attempts, human
458
+ nudges, changed files, untracked files, change-profile splits for project files versus AI wiki
459
+ telemetry and user-owned wiki churn, and whether the run is ready for shareable causal claims.
460
+ It does not run agents or mutate eval artifacts.
461
+
462
+ Use `eval impact summarize` to aggregate multiple captured runs into a product-level dashboard.
463
+ It reports each family's primary outcome, product signal, shareability, success and score deltas,
464
+ and change-profile deltas so neutral success-rate runs can still surface quality or churn signals.
465
+
436
466
  To diagnose missing starter pointers, stale managed prompt blocks, or rule drift and print copy-paste upgrade starters:
437
467
 
438
468
  ```bash
Binary file
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "ai-wiki-toolkit-linux-arm64",
3
- "version": "0.1.26",
3
+ "version": "0.1.28",
4
4
  "description": "Platform binary package for ai-wiki-toolkit (linux-arm64-glibc).",
5
5
  "license": "MIT",
6
6
  "homepage": "https://github.com/BochengYin/ai-wiki-toolkit#readme",