You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
@@ -256,7 +256,7 @@ The source tree can also be used directly for quick CLI inspection:
256
256
257
257
### Current main commands from source
258
258
259
-
Current `main` can also be tested from the source tree. `doctor`, `budget`, and `explain` are part of the published v0.3.0 package. `dedupe`and `conflicts` are post-v0.3.0 current-main additions until the next release is cut and verified:
259
+
Current `main` can also be tested from the source tree. `doctor`, `budget`, `explain`, `dedupe`, and `conflicts` are part of the published v0.4.0 packageand current source tree:
260
260
261
261
PYTHONPATH=src python -m agent_rules_kit.cli doctor tests/fixtures/repositories/multi-agent-overlap
These commands are implemented on current `main`. `doctor`, `budget`, and `explain`are part of the published v0.3.0 command surface. `dedupe` and `conflicts` are post-v0.3.0 `main` additions until the next release is cut and verified.
395
+
These commands are implemented on current `main`and are part of the published v0.4.0 command surface.
396
396
397
397
---
398
398
@@ -452,13 +452,17 @@ See:
452
452
│ ├── codeql.yml
453
453
│ └── publish-pypi.yml
454
454
├── docs/
455
+
│ ├── ACTIONS-PINNING-POLICY.md
455
456
│ ├── BUILD-PLAN.md
456
457
│ ├── DEPENDABOT-DEPENDENCY-GRAPH.md
457
458
│ ├── EXIT-CODES.md
459
+
│ ├── EXTERNAL-AUDIT-PACKAGE-CURRENT-MAIN.md
458
460
│ ├── OPENSSF-SCORECARD-EVALUATION.md
459
461
│ ├── OUTPUTS.md
462
+
│ ├── POST-AUDIT-ACTION-PLAN-CURRENT-MAIN.md
460
463
│ ├── POST-V0.3.0-FUNCTIONAL-CONTRACT-EVIDENCE.md
461
464
│ ├── POST-V0.3.0-INTERNAL-READINESS-AUDIT.md
465
+
│ ├── PRE-V0.4.0-INTERNAL-READINESS-AUDIT.md
462
466
│ ├── PRIVATE-VULNERABILITY-REPORTING.md
463
467
│ ├── PRODUCT-STRATEGY.md
464
468
│ ├── RULES.md
@@ -472,8 +476,14 @@ See:
472
476
│ ├── V0.3-ARCHITECTURE-ROADMAP.md
473
477
│ ├── V0.3.0-POST-RELEASE-AUDIT.md
474
478
│ ├── V0.3.0-RELEASE-NOTES.md
479
+
│ ├── V0.4.0-RELEASE-EVIDENCE.md
480
+
│ ├── V0.4.0-RELEASE-NOTES.md
475
481
│ └── screenshots/
476
482
│ └── readme/
483
+
│ ├── agent-rules-kit-governance-findings.png
484
+
│ ├── agent-rules-kit-help-check.png
485
+
│ ├── agent-rules-kit-init-safety.png
486
+
│ └── agent-rules-kit-output-formats.png
477
487
├── scripts/
478
488
│ ├── check.sh
479
489
│ └── post-release-audit.sh
@@ -492,6 +502,8 @@ See:
492
502
│ ├── init_write.py
493
503
│ └── redaction.py
494
504
├── tests/
505
+
│ ├── fixtures/
506
+
│ │ └── repositories/
495
507
│ ├── test_cli.py
496
508
│ ├── test_conflicts.py
497
509
│ ├── test_dedupe.py
@@ -543,7 +555,7 @@ Current verified local result on `main`:
543
555
544
556
The exact unit test count may change as coverage evolves. The source of truth is the current `./scripts/check.sh` output and the matching GitHub Actions run for `main`.
545
557
546
-
For current post-v0.3.0 functional evidence, including the verified command matrix, init write behavior, and release-boundary limits, see `docs/POST-V0.3.0-FUNCTIONAL-CONTRACT-EVIDENCE.md`.
558
+
For v0.4.0 release notes and release evidence, see `docs/V0.4.0-RELEASE-NOTES.md` and `docs/V0.4.0-RELEASE-EVIDENCE.md`.
547
559
548
560
CI installs project development dependencies and then runs the same local check script through GitHub Actions.
549
561
@@ -557,19 +569,19 @@ The required status check for `main` is:
557
569
558
570
Current status:
559
571
560
-
-`v0.3.0` is published as the current GitHub Release and PyPI package;
561
-
-`v0.2.3` remains the previous published GitHub Release and PyPI package baseline;
572
+
-`v0.4.0` is published as the current GitHub Release and PyPI package;
573
+
-`v0.3.0` remains the previous published GitHub Release and PyPI package baseline;
562
574
- no stable support or API guarantee yet;
563
-
- release tag `v0.3.0` points to the verified release SHA;
575
+
- release tag `v0.4.0` points to the verified release SHA;
564
576
- local CLI behavior implemented;
565
577
- governance diagnostics, structured finding evidence, and evidence redaction are implemented;
566
-
-`doctor`, `budget`, and `explain` are implemented as v0.3.0 commands, while `dedupe` and `conflicts` are implemented on current `main`as post-v0.3.0 read-only command additions;
578
+
-`doctor`, `budget`, `explain`, `dedupe`, and `conflicts` are implemented as v0.4.0 read-only command surfaces;
567
579
- CI active;
568
580
- branch protection is active with the required `local-checks / Python 3.12` status check;
569
581
- the `pypi` GitHub environment exists for the release publishing workflow;
570
-
-`.github/workflows/publish-pypi.yml` published `v0.3.0` through PyPI Trusted Publishing and remains the release publishing workflow;
582
+
-`.github/workflows/publish-pypi.yml` published `v0.4.0` through PyPI Trusted Publishing and remains the release publishing workflow;
571
583
- README screenshots are generated from real local CLI commands;
572
-
-post-v0.3.0 functional contract evidence is documented in `docs/POST-V0.3.0-FUNCTIONAL-CONTRACT-EVIDENCE.md`;
584
+
- v0.4.0 release notes and release evidence are documented in `docs/V0.4.0-RELEASE-NOTES.md` and `docs/V0.4.0-RELEASE-EVIDENCE.md`;
Copy file name to clipboardExpand all lines: SECURITY.md
+7-6Lines changed: 7 additions & 6 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -6,20 +6,21 @@ It is not a security scanner, provides no security guarantees, and must not be d
6
6
7
7
## Supported versions
8
8
9
-
`v0.3.0` is the current published GitHub Release and PyPI package.
9
+
`v0.4.0` is the current published GitHub Release and PyPI package.
10
10
11
-
`v0.2.3` remains the previous published GitHub Release and PyPI package baseline.
11
+
`v0.3.0` remains the previous published GitHub Release and PyPI package baseline.
12
12
13
13
The project is still maintained on a best-effort basis. There is no commercial SLA, no guaranteed response time, and no guarantee that every security-relevant issue will be found or fixed.
14
14
15
15
| Version | Status |
16
16
| --- | --- |
17
-
| 0.3.x | Current published GitHub Release line / best-effort security fixes |
18
-
| 0.2.x | Previous published line / limited best-effort fixes |
17
+
| 0.4.x | Current published GitHub Release line / best-effort security fixes |
18
+
| 0.3.x | Previous published line / limited best-effort fixes |
19
+
| 0.2.x | Historical line / limited best-effort fixes |
19
20
| 0.1.x | Historical pre-release line / not supported |
20
21
| < 0.1 | Not supported |
21
22
22
-
`agent-rules-kit==0.3.0` is published on PyPI. Future PyPI availability claims must be verified per release before updating this policy.
23
+
`agent-rules-kit==0.4.0` is published on PyPI. Future PyPI availability claims must be verified per release before updating this policy.
23
24
24
25
## Security boundaries
25
26
@@ -70,7 +71,7 @@ agent-rules-kit does not aim to:
70
71
71
72
## Maintainer response
72
73
73
-
Security response is best-effort for the current published `0.3.x` GitHub Release line.
74
+
Security response is best-effort for the current published `0.4.x` GitHub Release line.
74
75
75
76
There is no commercial SLA or guaranteed response time.
Copy file name to clipboardExpand all lines: docs/OUTPUTS.md
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -18,7 +18,7 @@ Implemented command surface:
18
18
-`agent-rules-kit conflicts`;
19
19
-`agent-rules-kit explain`.
20
20
21
-
`doctor`, `budget`, and `explain` are implemented as v0.3.0 command baselines. `dedupe` and `conflicts` are implemented on current `main` after v0.3.0 and must not be described as published package behavior until the next release is cut and verified. Release publication is verified by the dedicated GitHub Release and PyPI workflow evidence, not by this output contract alone.
21
+
The full command surface listed above is implemented in the published v0.4.0 command baseline. Release publication is verified by the dedicated GitHub Release and PyPI workflow evidence, not by this output contract alone.
Copy file name to clipboardExpand all lines: docs/PRODUCT-STRATEGY.md
+5-7Lines changed: 5 additions & 7 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -1,12 +1,12 @@
1
1
# Product Strategy Roadmap
2
2
3
-
This document defines the product direction for agent-rules-kit after the published v0.3.0 GitHub Release and PyPI package, with current post-v0.3.0 maintenance hardening on main.
3
+
This document defines the product direction for agent-rules-kit after the published v0.4.0 GitHub Release and PyPI package.
4
4
5
5
It is a strategy document, not an implementation plan for a specific feature. It must not be used to claim capabilities that are not already implemented.
6
6
7
7
## Current product truth
8
8
9
-
agent-rules-kit has a published v0.3.0 GitHub Release and PyPI package line. `v0.2.3` remains the previous published GitHub Release and PyPI package baseline. Current main contains post-v0.3.0 `dedupe` and `conflicts` command additions plus maintenance hardening. The next public release target is v0.4.0, not a v0.3.x patch, because the new commands expand the compatible command surface.
9
+
agent-rules-kit has a published v0.4.0 GitHub Release and PyPI package line. `v0.3.0` remains the previous published GitHub Release and PyPI package baseline. The v0.4.0 release publishes the compatible `dedupe` and `conflicts` command additions plus maintenance hardening that landed after v0.3.0.
10
10
11
11
The implemented product currently supports:
12
12
@@ -28,7 +28,6 @@ The implemented product does not yet provide:
28
28
29
29
- governance scoring;
30
30
- cross-file consistency analysis;
31
-
- instruction conflict detection;
32
31
- exact tokenizer-specific token accounting;
33
32
- policy profiles;
34
33
- remediation workflows;
@@ -266,13 +265,12 @@ Candidate criteria:
266
265
- private vulnerability reporting path or explicit documented alternative;
267
266
- no unsupported security or maturity claims.
268
267
269
-
## Next implementation direction after v0.3.0
268
+
## Next implementation direction after v0.4.0
270
269
271
270
The next implementation phase should be justified against the current repository state, not against older strategy snapshots.
272
271
273
272
Good next candidates are narrow and evidence-backed:
274
273
275
-
- v0.4.0 release preparation for the current `dedupe` and `conflicts` command additions, after final audit evidence is complete;
276
274
- release and package smoke hardening;
277
275
- README, support, security, and strategy public-truth checks;
278
276
- supply-chain workflow additions only after dedicated evaluation phases;
@@ -287,8 +285,8 @@ Decision: keep product strategy as the boundary document after v0.2.
287
285
288
286
Reason:
289
287
290
-
- v0.3.0 already published the local diagnosis toolkit baseline;
291
-
-current main contains post-v0.3.0 `dedupe` and `conflicts` command additions plus maintenance hardening that must be released, if approved, as v0.4.0 rather than a patch release;
288
+
- v0.4.0 publishes the current local instruction-governance command baseline;
289
+
-future post-v0.4.0 work should avoid broadening the tool before machine-readable output and renderer structure are improved;
292
290
- adjacent tools still cover repository packaging, context frameworks, and broad rule generation better than this project should;
293
291
- the real product wedge remains instruction governance;
294
292
- future features must be justified against this document to avoid scope drift.
Copy file name to clipboardExpand all lines: docs/THREAT-MODEL.md
+5-5Lines changed: 5 additions & 5 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -1,6 +1,6 @@
1
1
# Threat Model
2
2
3
-
This document defines the threat model for agent-rules-kit. It was last reviewed for the published v0.3.0 release line and the current post-v0.3.0 main state, including the published v0.3.0 doctor, budget, and explain command surface and the current-main post-v0.3.0 dedupe and conflicts read-only command additions.
3
+
This document defines the threat model for agent-rules-kit. It was last reviewed for the published v0.4.0 release line, including the read-only check, init, doctor, budget, explain, dedupe, and conflicts command surface.
4
4
5
5
It is intentionally narrow. agent-rules-kit is a local CLI for diagnosing AI agent instruction files. It is not a security scanner, malware detector, CI/CD auditor, sandbox, secret manager, or autonomous remediation agent.
6
6
@@ -14,13 +14,13 @@ In scope for the current release line and current main state:
14
14
- read-only doctor repository diagnosis summaries;
15
15
- read-only budget size and context-pressure approximation output;
16
16
- read-only explain output for local governance rule IDs;
17
-
- read-only dedupe duplicate instruction-line detection on current main as a post-v0.3.0 addition;
18
-
- read-only conflicts contradictory-guidance detection on current main as a post-v0.3.0 addition;
- explicit init write behavior for baseline AGENTS.md creation or replacement with backup;
21
21
- secret-like value redaction in findings and output.
22
22
23
-
`dedupe` and `conflicts` are current-main post-v0.3.0 additions. They must not be represented as published PyPI package behavior until a later release is cut, published, and verified.
23
+
`dedupe` and `conflicts` are included in the published v0.4.0 command surface after release publication and PyPI verification.
24
24
25
25
Out of scope:
26
26
@@ -163,7 +163,7 @@ Before release, verify:
163
163
- backups are created before replacing existing AGENTS.md;
164
164
- no runtime network or LLM dependency was introduced;
165
165
- no repository command execution was introduced;
166
-
-post-v0.3.0 `dedupe` and `conflicts` command boundaries remain read-only if they are included in the next release scope;
166
+
-`dedupe` and `conflicts` command boundaries remain read-only in the v0.4.0 release scope;
167
167
- secret-like examples in tests and docs are fake or redacted;
0 commit comments