agent-contracts 0.18.0 → 0.19.0

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.
@@ -2,7 +2,7 @@
2
2
 
3
3
  Declarative YAML DSL toolkit for defining, validating, and rendering multi-agent development workflows. Provides static validation, semantic linting, prompt rendering, guardrail generation, and completeness scoring for agent contract definitions.
4
4
 
5
- **Version:** 0.16.2
5
+ **Version:** 0.19.0
6
6
 
7
7
  ## Table of Contents
8
8
 
@@ -13,6 +13,7 @@ Declarative YAML DSL toolkit for defining, validating, and rendering multi-agent
13
13
  - [render](#agent-contracts-render)
14
14
  - [check](#agent-contracts-check)
15
15
  - [score](#agent-contracts-score)
16
+ - [audit](#agent-contracts-audit)
16
17
  - [generate](#agent-contracts-generate)
17
18
 
18
19
  ---
@@ -378,6 +379,89 @@ x-agent:
378
379
 
379
380
  ---
380
381
 
382
+ ### audit
383
+
384
+ Run LLM-based semantic audit on DSL definitions and generated outputs.
385
+
386
+ Performs LLM-based semantic analysis on DSL definitions and/or generated outputs. Three audit types are available: "render" checks whether generated files faithfully reflect the resolved DSL, "dsl" reviews DSL design for semantic coherence, and "prompt" verifies that generated prompts accurately express DSL intent. Uses agent-contracts-runtime (optional peer dependency) for LLM execution with handoff schema validation and follow-up recovery.
387
+
388
+ **Usage:**
389
+
390
+ ```
391
+ agent-contracts audit render -c agent-contracts.config.yaml
392
+ ```
393
+ ```
394
+ agent-contracts audit dsl -c agent-contracts.config.yaml
395
+ ```
396
+ ```
397
+ agent-contracts audit prompt -c agent-contracts.config.yaml
398
+ ```
399
+ ```
400
+ agent-contracts audit all -c agent-contracts.config.yaml
401
+ ```
402
+ ```
403
+ agent-contracts audit dsl --dry-run -c agent-contracts.config.yaml
404
+ ```
405
+ ```
406
+ agent-contracts audit render --format json -c agent-contracts.config.yaml
407
+ ```
408
+ ```
409
+ agent-contracts audit dsl --scope agents:architect,implementer -c config.yaml
410
+ ```
411
+
412
+ #### Arguments
413
+
414
+ | Name | Required | Description |
415
+ |---|---|---|
416
+ | `type` | No | Audit type to run: render (semantic diff of generated outputs vs DSL), dsl (design coherence review), prompt (generated prompt fidelity check), or all (run all three). |
417
+
418
+ #### Options
419
+
420
+ | Option | Aliases | Required | Default | Description |
421
+ |---|---|---|---|---|
422
+ | `--config` | -c | No | | Path to agent-contracts.config.yaml. |
423
+ | `--team` | | No | | Limit to one team (multi-team config only). |
424
+ | `--format` | | No | `"text"` | Output format. |
425
+ | `--scope` | | No | | Limit audit scope to specified entities (e.g. agents:architect,implementer). |
426
+ | `--dry-run` | | No | `false` | Output the audit prompt without calling LLM. |
427
+ | `--adapter` | | No | | SDK adapter to use for LLM calls (overrides config audit.adapter). |
428
+ | `--model` | | No | | LLM model override (overrides config audit.model). |
429
+
430
+ #### Exit Codes
431
+
432
+ **Exit 0:** No critical findings detected.
433
+
434
+ - **stdout:** format=`text`
435
+
436
+ **Exit 1:** Critical findings detected.
437
+
438
+ - **stdout:** format=`text`
439
+
440
+ **Exit 2:** Invalid input or configuration error.
441
+
442
+ - **stderr:** format=`text`
443
+
444
+ **Exit 3:** LLM adapter error (API failure, runtime not installed).
445
+
446
+ - **stderr:** format=`text`
447
+
448
+ #### Extensions
449
+
450
+ ```yaml
451
+ x-agent:
452
+ riskLevel: low
453
+ requiresConfirmation: false
454
+ idempotent: true
455
+ sideEffects:
456
+
457
+ recommendedBeforeUse:
458
+ - Ensure agent-contracts.config.yaml exists with render targets.
459
+ - Run validate first to confirm DSL is valid.
460
+ - Install agent-contracts-runtime if not using --dry-run.
461
+ ```
462
+
463
+ ---
464
+
381
465
  ### generate
382
466
 
383
467
  Generate guardrail or interface artifacts from DSL.
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "agent-contracts",
3
- "version": "0.18.0",
3
+ "version": "0.19.0",
4
4
  "description": "Declarative YAML DSL toolkit for defining, validating, and rendering multi-agent development workflows",
5
5
  "type": "module",
6
6
  "main": "dist/index.js",
@@ -69,11 +69,16 @@
69
69
  "zod-to-json-schema": "^3.25.2"
70
70
  },
71
71
  "devDependencies": {
72
+ "@anthropic-ai/claude-agent-sdk": "^0.2.138",
73
+ "@cursor/sdk": "^1.0.12",
72
74
  "@eslint/js": "^10.0.1",
75
+ "@google/genai": "^2.0.1",
76
+ "@openai/agents": "^0.11.1",
73
77
  "@types/node": "^25.6.0",
78
+ "agent-contracts-runtime": "^0.12.0",
79
+ "cli-contracts": "^0.2.0",
74
80
  "eslint": "^10.2.0",
75
81
  "globals": "^17.5.0",
76
- "cli-contracts": "^0.2.0",
77
82
  "prettier": "^3.8.2",
78
83
  "tsup": "^8.5.1",
79
84
  "tsx": "^4.21.0",
@@ -81,6 +86,11 @@
81
86
  "typescript-eslint": "^8.58.1",
82
87
  "vitest": "^4.1.4"
83
88
  },
89
+ "peerDependenciesMeta": {
90
+ "agent-contracts-runtime": {
91
+ "optional": true
92
+ }
93
+ },
84
94
  "engines": {
85
95
  "node": ">=20.0.0"
86
96
  }