sanityops-cli 0.1.9__tar.gz → 0.1.10__tar.gz

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.
Files changed (135) hide show
  1. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/CHANGELOG.md +2 -2
  2. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/PKG-INFO +1 -18
  3. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/README.md +0 -17
  4. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/plans/2026-09-08-config-command.md +1 -1
  5. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/plans/2026-09-20-cross-artifact-rendering.md +1 -1
  6. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/plans/2026-09-20-help-panel-migration.md +2 -2
  7. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/plans/2026-09-20-repair-feature.md +2 -2
  8. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/specs/2026-09-08-config-command-design.md +2 -2
  9. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/specs/2026-09-20-cross-artifact-rendering-design.md +1 -1
  10. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/specs/2026-09-20-help-panel-migration-design.md +6 -6
  11. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/specs/2026-09-20-repair-feature-design.md +4 -4
  12. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/pyproject.toml +1 -1
  13. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/__init__.py +1 -1
  14. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/agents/scanner_agent/models/finding.py +36 -7
  15. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/agents/scanner_agent/prompts.py +114 -18
  16. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/agents/scanner_agent/tools/grep_tool.py +1 -1
  17. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/agents/scanner_agent/tools/readfile_tool.py +1 -1
  18. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/commands/inspect.py +1 -1
  19. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/defect_checker/markdown_reporter.py +3 -27
  20. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/defect_checker/renderer.py +3 -13
  21. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/help_panel.py +2 -3
  22. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/commands/test_inspect_repair.py +8 -9
  23. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/test_help_panel.py +1 -3
  24. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/defect_checker/test_markdown_reporter.py +79 -24
  25. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/.github/ISSUE_TEMPLATE/bug_report.yml +0 -0
  26. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/.github/ISSUE_TEMPLATE/config.yml +0 -0
  27. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/.github/ISSUE_TEMPLATE/feature_request.yml +0 -0
  28. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/.github/PULL_REQUEST_TEMPLATE.md +0 -0
  29. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/.github/dependabot.yml +0 -0
  30. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/.github/scripts/pr_review_bot.py +0 -0
  31. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/.github/workflows/ai-review.yml +0 -0
  32. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/.github/workflows/ci.yml +0 -0
  33. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/.github/workflows/release.yml +0 -0
  34. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/.gitignore +0 -0
  35. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/CODE_OF_CONDUCT.md +0 -0
  36. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/CONTRIBUTING.md +0 -0
  37. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/LICENSE +0 -0
  38. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/NOTICE +0 -0
  39. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/SECURITY.md +0 -0
  40. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/plans/2026-08-13-model-config-comments-in-init-template.md +0 -0
  41. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/plans/2026-08-14-loading-animations.md +0 -0
  42. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/plans/2026-08-14-skills-files-banner-readme.md +0 -0
  43. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/plans/2026-08-28-copyright-compliance.md +0 -0
  44. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/specs/2026-08-13-model-config-comments-in-init-template-design.md +0 -0
  45. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/specs/2026-08-14-loading-animations-design.md +0 -0
  46. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/specs/2026-08-14-skills-files-banner-readme-design.md +0 -0
  47. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/docs/superpowers/specs/2026-08-28-copyright-compliance-design.md +0 -0
  48. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/entry.py +0 -0
  49. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/release.sh +0 -0
  50. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/scripts/generate-checksums.ps1 +0 -0
  51. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/scripts/install.ps1 +0 -0
  52. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/scripts/install.sh +0 -0
  53. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/agents/__init__.py +0 -0
  54. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/agents/repair_agent/__init__.py +0 -0
  55. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/agents/repair_agent/agent.py +0 -0
  56. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/agents/repair_agent/prompts.py +0 -0
  57. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/agents/repair_agent/tools/__init__.py +0 -0
  58. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/agents/repair_agent/tools/store_repairs_tool.py +0 -0
  59. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/agents/scanner_agent/agent.py +0 -0
  60. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/agents/scanner_agent/hooks/progress_hook.py +0 -0
  61. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/agents/scanner_agent/tools/listfiles_tool.py +0 -0
  62. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/agents/scanner_agent/tools/storefindings_tool.py +0 -0
  63. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/api/__init__.py +0 -0
  64. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/api/client.py +0 -0
  65. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/code_contract_checker/__init__.py +0 -0
  66. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/code_contract_checker/checker.py +0 -0
  67. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/code_contract_checker/markdown_reporter.py +0 -0
  68. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/code_contract_checker/models.py +0 -0
  69. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/code_contract_checker/renderer.py +0 -0
  70. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/code_contract_checker/rules/__init__.py +0 -0
  71. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/code_contract_checker/rules/qc0_baseline.py +0 -0
  72. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/code_contract_checker/rules/qc1_syntax.py +0 -0
  73. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/code_contract_checker/rules/qc2_resource.py +0 -0
  74. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/code_contract_checker/rules/qc3_pattern.py +0 -0
  75. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/commands/__init__.py +0 -0
  76. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/commands/config.py +0 -0
  77. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/commands/init.py +0 -0
  78. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/constants/__init__.py +0 -0
  79. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/constants/config_defaults.py +0 -0
  80. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/constants/exit_codes.py +0 -0
  81. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/defect_checker/__init__.py +0 -0
  82. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/defect_checker/checker.py +0 -0
  83. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/defect_checker/llm_config.py +0 -0
  84. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/exceptions/__init__.py +0 -0
  85. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/exceptions/api_exceptions.py +0 -0
  86. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/exceptions/base_exceptions.py +0 -0
  87. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/logging/__init__.py +0 -0
  88. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/logging/logger.py +0 -0
  89. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/main.py +0 -0
  90. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/progress/__init__.py +0 -0
  91. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/progress/tracker.py +0 -0
  92. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/renderers/__init__.py +0 -0
  93. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/renderers/command_renderer/inspect_command_renderer.py +0 -0
  94. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/templates/__init__.py +0 -0
  95. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/templates/inspect_config.yaml +0 -0
  96. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/utils/__init__.py +0 -0
  97. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/utils/artifact_packer.py +0 -0
  98. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/utils/config_loader.py +0 -0
  99. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/utils/config_resolver.py +0 -0
  100. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/src/sanityops_cli/utils/validators.py +0 -0
  101. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/__init__.py +0 -0
  102. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/commands/__init__.py +0 -0
  103. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/commands/test_config.py +0 -0
  104. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/commands/test_init.py +0 -0
  105. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/__init__.py +0 -0
  106. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/agents/__init__.py +0 -0
  107. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/agents/repair_agent/__init__.py +0 -0
  108. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/agents/repair_agent/test_agent.py +0 -0
  109. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/agents/repair_agent/test_store_repairs_tool.py +0 -0
  110. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/agents/scanner_agent/__init__.py +0 -0
  111. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/agents/scanner_agent/models/__init__.py +0 -0
  112. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/agents/scanner_agent/models/test_finding.py +0 -0
  113. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/agents/scanner_agent/test_agent.py +0 -0
  114. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/agents/scanner_agent/test_deterministic_capture.py +0 -0
  115. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/agents/scanner_agent/test_progress_hook.py +0 -0
  116. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/agents/scanner_agent/test_prompts.py +0 -0
  117. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/api/__init__.py +0 -0
  118. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/api/test_client.py +0 -0
  119. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/code_contract_checker/__init__.py +0 -0
  120. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/code_contract_checker/test_checker.py +0 -0
  121. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/code_contract_checker/test_renderer.py +0 -0
  122. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/code_contract_checker/test_rules.py +0 -0
  123. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/commands/test_inspect.py +0 -0
  124. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/commands/test_version.py +0 -0
  125. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/defect_checker/__init__.py +0 -0
  126. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/defect_checker/test_checker.py +0 -0
  127. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/defect_checker/test_llm_config.py +0 -0
  128. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/defect_checker/test_renderer.py +0 -0
  129. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/logging/__init__.py +0 -0
  130. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/logging/test_logger.py +0 -0
  131. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/progress/__init__.py +0 -0
  132. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/progress/test_tracker.py +0 -0
  133. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/utils/test_artifact_packer.py +0 -0
  134. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/unit/utils/test_config_loader.py +0 -0
  135. {sanityops_cli-0.1.9 → sanityops_cli-0.1.10}/tests/utils/test_config_resolver.py +0 -0
@@ -33,13 +33,13 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
33
33
 
34
34
  - `inspect repair` subcommand to generate repaired artifacts from inspection reports
35
35
  - Cross-artifact result rendering with aggregate CROSS score in markdown reports
36
- - Getting Started and Advanced Usage help panels migrated from deeplogic-cli
36
+ - Getting Started and Advanced Usage help panels migrated from sanityops-cli
37
37
 
38
38
  ## [0.0.7] - 2026-09-10
39
39
 
40
40
  ### Added
41
41
 
42
- - Init command alignment with deeplogic-cli: placeholder UUID, project.name, model block
42
+ - Init command alignment with sanityops-cli: placeholder UUID, project.name, model block
43
43
  - Nested `.gitignore` in `.sanityops` directory to prevent accidental secret commits
44
44
 
45
45
  ### Changed
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.5
2
2
  Name: sanityops-cli
3
- Version: 0.1.9
3
+ Version: 0.1.10
4
4
  Summary: CLI tool for static defect inspection of AI logical artifacts
5
5
  Project-URL: Homepage, https://sanityops.org
6
6
  Project-URL: Documentation, https://sanityops.org
@@ -160,23 +160,6 @@ sanityops-cli inspect --config path/to/config.yaml
160
160
 
161
161
  `inspect` exits `0` when the run completes — including when defects are found — and non-zero on failure. CI pipelines should therefore gate on the generated report (`.sanityops/results/inspect-*.md`), not on the exit code.
162
162
 
163
- ### 4. Generate Repairs (Optional)
164
-
165
- After inspection, generate repaired artifacts from the report:
166
-
167
- ```bash
168
- # Use the latest inspection report
169
- sanityops-cli inspect repair
170
-
171
- # Use a specific report
172
- sanityops-cli inspect repair --report .sanityops/results/inspect-20250115-120000.md
173
-
174
- # Specify timeout and token budget for large projects
175
- sanityops-cli inspect repair --timeout 3600 --token-budget 500000
176
- ```
177
-
178
- Repairs are written to `.sanityops/repairs/repair-<timestamp>.md` without modifying source files.
179
-
180
163
  ## Data Handling
181
164
 
182
165
  What leaves your machine, and when:
@@ -120,23 +120,6 @@ sanityops-cli inspect --config path/to/config.yaml
120
120
 
121
121
  `inspect` exits `0` when the run completes — including when defects are found — and non-zero on failure. CI pipelines should therefore gate on the generated report (`.sanityops/results/inspect-*.md`), not on the exit code.
122
122
 
123
- ### 4. Generate Repairs (Optional)
124
-
125
- After inspection, generate repaired artifacts from the report:
126
-
127
- ```bash
128
- # Use the latest inspection report
129
- sanityops-cli inspect repair
130
-
131
- # Use a specific report
132
- sanityops-cli inspect repair --report .sanityops/results/inspect-20250115-120000.md
133
-
134
- # Specify timeout and token budget for large projects
135
- sanityops-cli inspect repair --timeout 3600 --token-budget 500000
136
- ```
137
-
138
- Repairs are written to `.sanityops/repairs/repair-<timestamp>.md` without modifying source files.
139
-
140
123
  ## Data Handling
141
124
 
142
125
  What leaves your machine, and when:
@@ -2,7 +2,7 @@
2
2
 
3
3
  > **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.
4
4
 
5
- **Goal:** Migrate the `config` subcommand from deeplogic-cli to sanityops-cli, providing CLI-based configuration management for server URL and API key.
5
+ **Goal:** Migrate the `config` subcommand from sanityops-cli to sanityops-cli, providing CLI-based configuration management for server URL and API key.
6
6
 
7
7
  **Architecture:** Create three new files (`constants/config_defaults.py`, `utils/config_resolver.py`, `commands/config.py`) and modify `main.py` to register the config subcommand. ConfigResolver handles the precedence chain (project > global > env > default) with YAML read/write. The config command provides get/set/list/unset operations via Typer CLI.
8
8
 
@@ -2,7 +2,7 @@
2
2
 
3
3
  > **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.
4
4
 
5
- **Goal:** Migrate cross-artifact result rendering from deeplogic-cli to sanityops-cli, enabling terminal and markdown display of CROSS module defects.
5
+ **Goal:** Migrate cross-artifact result rendering from sanityops-cli to sanityops-cli, enabling terminal and markdown display of CROSS module defects.
6
6
 
7
7
  **Architecture:** Add helper functions to resolve artifact names from defect references and aggregate multiple CROSS sub-results into a unified section. The terminal renderer uses Rich panels with artifact names in titles; the markdown reporter merges PS/PT/ST sub-results and calculates a unified score via SDK.
8
8
 
@@ -2,7 +2,7 @@
2
2
 
3
3
  > **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.
4
4
 
5
- **Goal:** Migrate Getting Started and Advanced Usage help panels from deeplogic-cli to sanityops-cli.
5
+ **Goal:** Migrate Getting Started and Advanced Usage help panels from sanityops-cli to sanityops-cli.
6
6
 
7
7
  **Architecture:** Copy `help_panel.py` verbatim, add `--help` interception logic in `main.py`, update entry points (`entry.py`, `pyproject.toml`) to call `main()` instead of `app()`, and add tests.
8
8
 
@@ -82,7 +82,7 @@ def get_advanced_usage_panel() -> Panel:
82
82
  )
83
83
  ```
84
84
 
85
- **Note:** Step 5 (`inspect cover`) from deeplogic-cli was removed because sanityops-cli has no `cover` subcommand.
85
+ **Note:** Step 5 (`inspect cover`) from sanityops-cli was removed because sanityops-cli has no `cover` subcommand.
86
86
 
87
87
  - [ ] **Step 2: Verify file is importable**
88
88
 
@@ -2,7 +2,7 @@
2
2
 
3
3
  > **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.
4
4
 
5
- **Goal:** Migrate local repair functionality from deeplogic-cli to sanityops-cli, enabling users to generate repaired artifact content from inspection reports.
5
+ **Goal:** Migrate local repair functionality from sanityops-cli to sanityops-cli, enabling users to generate repaired artifact content from inspection reports.
6
6
 
7
7
  **Architecture:** Create a RepairAgent that reads inspection reports and uses an LLM to rewrite defective artifacts. The agent uses StoreRepairsTool to collect repairs, which are then written to a markdown report file.
8
8
 
@@ -1438,5 +1438,5 @@ Create a test scenario:
1438
1438
 
1439
1439
  ```bash
1440
1440
  git add -A
1441
- git commit -m "feat(repair): complete repair feature migration from deeplogic-cli"
1441
+ git commit -m "feat(repair): complete repair feature migration from sanityops-cli"
1442
1442
  ```
@@ -2,11 +2,11 @@
2
2
 
3
3
  **Date:** 2026-09-08
4
4
  **Status:** Approved
5
- **Source:** deeplogic-cli → sanityops-cli
5
+ **Source:** sanityops-cli → sanityops-cli
6
6
 
7
7
  ## Overview
8
8
 
9
- Migrate the `config` subcommand from deeplogic-cli to sanityops-cli. This provides CLI-based configuration management for the sanityops server backend URL and API key.
9
+ Migrate the `config` subcommand from sanityops-cli to sanityops-cli. This provides CLI-based configuration management for the sanityops server backend URL and API key.
10
10
 
11
11
  ## Architecture
12
12
 
@@ -8,7 +8,7 @@
8
8
 
9
9
  sanityops-cli currently skips rendering of cross-artifact defects (module type "CROSS") in both terminal and markdown outputs. The defect-check SDK returns cross-artifact analysis results, but the CLI explicitly skips them.
10
10
 
11
- This design migrates the cross-artifact rendering functionality from deeplogic-cli to sanityops-cli.
11
+ This design migrates the cross-artifact rendering functionality from sanityops-cli to sanityops-cli.
12
12
 
13
13
  ## Migration Scope
14
14
 
@@ -1,12 +1,12 @@
1
1
  # Help Panel Migration — Design Spec
2
2
 
3
3
  **Date:** 2026-09-20
4
- **Source:** `/Users/a1234/Documents/project/gitlab/deeplogic-cli`
4
+ **Source:** `/Users/a1234/Documents/project/gitlab/sanityops-cli`
5
5
  **Target:** `/Users/a1234/Documents/project/sanityops-cli`
6
6
 
7
7
  ## Summary
8
8
 
9
- Migrate the Getting Started and Advanced Usage help panels from deeplogic-cli to sanityops-cli. The panels appear when users run `sanityops-cli --help` or `sanityops-cli` with no arguments, providing a 4-step quick start workflow and CI/CD integration guidance.
9
+ Migrate the Getting Started and Advanced Usage help panels from sanityops-cli to sanityops-cli. The panels appear when users run `sanityops-cli --help` or `sanityops-cli` with no arguments, providing a 4-step quick start workflow and CI/CD integration guidance.
10
10
 
11
11
  **Note:** The original 5-step workflow was reduced to 4 steps because sanityops-cli has no `inspect cover` subcommand.
12
12
 
@@ -14,7 +14,7 @@ Migrate the Getting Started and Advanced Usage help panels from deeplogic-cli to
14
14
 
15
15
  | File | Action |
16
16
  |------|--------|
17
- | `src/sanityops_cli/help_panel.py` | Create (localized from deeplogic-cli) |
17
+ | `src/sanityops_cli/help_panel.py` | Create (localized from sanityops-cli) |
18
18
  | `src/sanityops_cli/main.py` | Modify (add help interception logic) |
19
19
  | `entry.py` | Modify (call `main()` instead of `app()`) |
20
20
  | `pyproject.toml` | Modify (entry point `:app` → `:main`) |
@@ -130,8 +130,8 @@ sanityops-cli = "sanityops_cli.main:main"
130
130
 
131
131
  ### 5. `tests/test_help_panel.py` (new)
132
132
 
133
- Copy from deeplogic-cli with these changes:
134
- - Import path `deeplogic_cli` → `sanityops_cli`
133
+ Copy from sanityops-cli with these changes:
134
+ - Import path `sanityops_cli` → `sanityops_cli`
135
135
  - **Remove** `test_getting_started_text_is_english()` and `test_advanced_usage_text_is_english()` (CJK assertion tests — will be added back when text is localized)
136
136
 
137
137
  Tests included:
@@ -169,7 +169,7 @@ sanityops-cli inspect --help
169
169
  ## Out of Scope
170
170
 
171
171
  - Adding `project` command (does not exist in sanityops-cli)
172
- - Migrating other deeplogic-cli features
172
+ - Migrating other sanityops-cli features
173
173
 
174
174
  ## Acceptance Criteria
175
175
 
@@ -2,11 +2,11 @@
2
2
 
3
3
  **Date:** 2026-09-20
4
4
  **Status:** Approved
5
- **Source:** Migration from deeplogic-cli
5
+ **Source:** Migration from sanityops-cli
6
6
 
7
7
  ## Overview
8
8
 
9
- Migrate the local `repair` feature from deeplogic-cli to sanityops-cli. The repair feature reads an inspection report and uses an LLM agent to generate fixed artifact content.
9
+ Migrate the local `repair` feature from sanityops-cli to sanityops-cli. The repair feature reads an inspection report and uses an LLM agent to generate fixed artifact content.
10
10
 
11
11
  ## Scope
12
12
 
@@ -241,8 +241,8 @@ Use the store_repairs tool with operation "add" for each repaired artifact:
241
241
 
242
242
  **Method:** Direct migration with import path adjustment
243
243
 
244
- 1. Copy `RepairAgent` from deeplogic-cli
245
- 2. Copy `StoreRepairsTool` from deeplogic-cli
244
+ 1. Copy `RepairAgent` from sanityops-cli
245
+ 2. Copy `StoreRepairsTool` from sanityops-cli
246
246
  3. Change imports: `aidynamic_agent` → `sanityops_agent`
247
247
  4. Add `repair` subcommand to `inspect.py`
248
248
  5. Reuse existing utilities:
@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
4
4
 
5
5
  [project]
6
6
  name = "sanityops-cli"
7
- version = "0.1.9"
7
+ version = "0.1.10"
8
8
  description = "CLI tool for static defect inspection of AI logical artifacts"
9
9
  readme = "README.md"
10
10
  requires-python = ">=3.11"
@@ -13,4 +13,4 @@
13
13
  # limitations under the License.
14
14
  #
15
15
 
16
- __version__ = "0.1.9"
16
+ __version__ = "0.1.10"
@@ -30,6 +30,7 @@ class FindingType(StrEnum):
30
30
 
31
31
  class Section(BaseModel):
32
32
  """A markdown section within a skill file."""
33
+
33
34
  title: str = Field(description="Section heading text")
34
35
  content: str = Field(description="Section body content")
35
36
 
@@ -37,7 +38,9 @@ class Section(BaseModel):
37
38
  class ToolSchema(BaseModel):
38
39
  name: str
39
40
  description: str = ""
40
- parameters: dict = Field(default_factory=dict,description="Tool parameters, JSON Schema (Draft-07) format")
41
+ parameters: dict = Field(
42
+ default_factory=dict, description="Tool parameters, JSON Schema (Draft-07) format"
43
+ )
41
44
 
42
45
 
43
46
  class SkillContent(BaseModel):
@@ -54,7 +57,11 @@ class SkillContent(BaseModel):
54
57
  #: Frontmatter fields that must be positive integers (QDS-0.4). LLM-based
55
58
  #: extraction can stringify numeric values, so coerce them back to int.
56
59
  _POSITIVE_INT_FIELDS: tuple[str, ...] = (
57
- "max_items", "max_chars", "timeout_seconds", "max_tool_calls", "max_retries",
60
+ "max_items",
61
+ "max_chars",
62
+ "timeout_seconds",
63
+ "max_tool_calls",
64
+ "max_retries",
58
65
  )
59
66
 
60
67
  @model_validator(mode="after")
@@ -100,23 +107,45 @@ class SkillContent(BaseModel):
100
107
  if not parts:
101
108
  return ""
102
109
  parts.append(
103
- "\n\n".join(
104
- f"## {section.title}\n\n{section.content}"
105
- for section in self.sections
106
- )
110
+ "\n\n".join(f"## {section.title}\n\n{section.content}" for section in self.sections)
107
111
  )
108
112
  return "\n".join(parts)
109
113
 
110
114
 
111
115
  class ToolContent(BaseModel):
112
116
  """Extracted content from a tool file."""
117
+
113
118
  name: str = Field(description="Tool name from definition")
114
119
  description: str = Field(description="Tool description")
115
- parameters: dict[str, Any] = Field(default_factory=dict, description="JSON Schema of parameters")
120
+ parameters: dict[str, Any] = Field(
121
+ default_factory=dict, description="JSON Schema of parameters"
122
+ )
123
+ #: pre-inspect-extract alignment: extraction fidelity labels. Default empty
124
+ #: when the analyzer does not provide them (backward compatible).
125
+ completeness: str = Field(
126
+ default="",
127
+ description=(
128
+ "Completeness level per pre-inspect-extract: full original / "
129
+ "assembled / incomplete / summary / not inspectable"
130
+ ),
131
+ )
132
+ evidence: str = Field(
133
+ default="", description="Short verbatim excerpt from the source proving the extraction"
134
+ )
116
135
 
117
136
 
118
137
  class PromptContent(BaseModel):
119
138
  content: str
139
+ #: pre-inspect-extract alignment: extraction fidelity labels. Default empty
140
+ #: when the analyzer does not provide them (backward compatible).
141
+ completeness: str = Field(
142
+ default="",
143
+ description=(
144
+ "Completeness level per pre-inspect-extract: full original / "
145
+ "original · variable unexpanded / assembled / incomplete / summary / not inspectable"
146
+ ),
147
+ )
148
+ evidence: str = Field(default="", description="Short verbatim excerpt from the source")
120
149
 
121
150
 
122
151
  class Finding(BaseModel):
@@ -21,15 +21,27 @@ A directory is a **Skill** if it contains:
21
21
  - `SKILL.md` file (primary indicator), OR
22
22
  - `README.md` with skill-defining frontmatter/headings (`# Skill`, `# Workflow`)
23
23
 
24
- ## 2. Exclusion Rules (STRICT)
24
+ A skill may also be **carried inside code** (e.g. a `@skill` decorator, a skill class/definition embedded in a `.py`/`.ts`/`.json` file). In that case treat it like a tool/prompt carrier: strip the code shell and extract the skill definition verbatim; classify the file by what it actually contains.
25
+
26
+ A skill is only inspectable when its definition carries the structural header — `name`/`description` frontmatter. A bare `## Skills` list or a bare function call that only *describes* a capability is documentation, not an artifact; skip it, do not inspect it.
27
+
28
+ **Four-way classification** (before extracting, decide what the candidate is):
29
+ - **Artifact** — the Skill definition itself (with structural header). Send to inspection.
30
+ - **Carrier** — code/markup that wraps a skill definition (decorator, class, embedded dict). Strip the shell, extract verbatim.
31
+ - **Scaffolding** — metadata, `Evaluation Dimensions`, `Scoring Criteria`, `Test Suggestions`, `Notes`. Discard.
32
+ - **Reference file** — pure code or pure data that contains no extractable skill definition. Skip it — do not report it as a skill.
33
+
34
+ ## 2. Exclusion Rules (STRICT) - HIGHEST PRIORITY
35
+ These directories MUST be excluded from all search operations:
25
36
  - Hidden paths starting with `.` (e.g., `.git`, `.github`, `.vscode`, `.idea`, `.venv`, `.env`, `.cache`)
26
37
  - Common dependency & build output folders (e.g., `node_modules`, `vendor`, `dist`, `build`, `out`, `__pycache__`, `.egg-info`, `site-packages`)
27
38
  - Test and fixture directories (e.g., `tests/`, `__tests__/`, `fixtures/`, `mocks/`)
28
39
  - Irrelevant non-skill docs (e.g., root `README.md`, `CONTRIBUTING.md`, `LICENSE`, `CHANGELOG.md`, component/API docs)
40
+ - Scaffolding content: `Evaluation Dimensions`, `Scoring Criteria`, `Test Suggestions`, `Notes` (metadata describing a skill, not the skill itself)
29
41
 
30
- ## 3. Search Strategy
31
- 1. Use `glob` to find `**/SKILL.md` within max depth `{depth}`
32
- 2. Verify README.md for skill markers
42
+ ## 3. Search Strategy (Exclusions applied at each step)
43
+ 1. Use `glob` to find `**/SKILL.md` within max depth `{depth}` - automatically exclude paths matching Section 2 rules
44
+ 2. Verify README.md for skill markers - skip excluded directories
33
45
  3. Confirm directory structure
34
46
 
35
47
  ## 4. Reporting via store_findings
@@ -53,6 +65,15 @@ Return summary when done: "Completed: found N skills in {directory}"
53
65
  TOOL_FINDER_RULES = """\
54
66
  You are an expert tool discovery agent. Find all **tools** in `{directory}` and report each one immediately via `store_findings`.
55
67
 
68
+ ## 0. Artifact vs Carrier vs Scaffolding vs Reference File (STRICT)
69
+ Before extracting a tool, classify what you see:
70
+ - **Artifact** — the Tool Schema itself (JSON Schema / function-calling definition with `name` + `description` + `parameters`/`inputSchema`). Send to inspection.
71
+ - **Carrier** — code or markup that wraps a tool definition (e.g. `def f(...): return ...`, a `@tool` decorator, a `Tool(...)` constructor call, backtick strings). A carrier is NOT a tool; strip the shell and extract the schema inside.
72
+ - **Scaffolding** — metadata, Evaluation Dimensions, Scoring Criteria, Test Suggestions, Notes, example usage, registration lists. Discard — never report scaffolding as a tool.
73
+ - **Reference file** — pure code or pure data that contains no extractable tool schema (e.g. a utility module, a data file). Skip it — do not report it as a tool.
74
+
75
+ Do NOT enumerate syntax case by case: `Tool(...)`, `@tool`, `BaseTool`, backticks and `"..."` differ only in which shell to strip, not in whether a tool is present.
76
+
56
77
  ## 1. Framework-Aware Search Patterns
57
78
 
58
79
  Identify tools by recognizing framework-specific patterns:
@@ -73,16 +94,45 @@ Identify tools by recognizing framework-specific patterns:
73
94
  - OpenAPI specs: `"type": "function"` with `operationId`
74
95
  - JSON/YAML tool configs: Anthropic `"input_schema"`, OpenAI `"parameters"`
75
96
 
76
- ## 2. Exclusion Rules (STRICT)
97
+ ## 1.5 Extraction Tiers (pre-inspect-extract alignment)
98
+ For each candidate tool, determine which tier it falls into and state the basis:
99
+
100
+ - **Tier 1 — directly readable**: the schema exists in loadable form — a complete
101
+ JSON Schema inside ``` fences, or a `Tool(name=..., description=..., inputSchema=<json-schema-dict>)`
102
+ where `inputSchema` is already a JSON Schema dict, or a pure JSON Schema.
103
+ → **verbatim copy**: take the schema as-is, drop only the surrounding shell.
104
+ - **Tier 2 — wrapped in code**: the schema is complete but carried inside code
105
+ (a dict literal in a function, a `.parameters`/`input_schema` object property).
106
+ → **verbatim copy**: extract the literal, unescape, keep variables as-is.
107
+ - **Tier 3 — only code, no ready-made schema**: constraints are written in
108
+ chained calls (Zod `z.object(...)`), decorators with type annotations,
109
+ or builder patterns (`mcp.NewTool(...)` + `WithString/...`).
110
+ → **assembly**: copy field-by-field everything already literal in the source
111
+ (`name`, `description`, each parameter key + description, `enum` values,
112
+ the `required` list, `default` values). **Add no field absent from the source.**
113
+ If a constraint cannot be converted (e.g. `.refine()`), do not fabricate an
114
+ equivalent — note it as not representable.
115
+ **The `description` field must not be lost**: description / `.describe()` /
116
+ docstring is the core basis of Tool Schema inspection and is the easiest thing
117
+ to lose during assembly. If the source has a description but the result does
118
+ not, label the whole thing `assembled` and record "description field at risk
119
+ of loss". If the source has no description, the result must not have one either
120
+ — do not write one.
121
+ - **Tier X — unrecognizable**: no structural signature, dynamic registration
122
+ with no literal schema, or only a registration list without `inputSchema`/`parameters`.
123
+ → label **not inspectable**, record the reason, and do not force a verdict.
124
+
125
+ ## 2. Exclusion Rules (STRICT) - HIGHEST PRIORITY
126
+ These directories MUST be excluded from all search operations:
77
127
  - Hidden paths starting with `.` (e.g., `.git`, `.github`, `.vscode`, `.idea`, `.venv`, `.env`, `.cache`)
78
128
  - Common dependency & build output folders (e.g., `node_modules`, `vendor`, `dist`, `build`, `out`, `__pycache__`, `.egg-info`, `site-packages`)
79
129
  - Test and fixture directories (e.g., `tests/`, `__tests__/`, `fixtures/`, `mocks/`)
80
130
  - Irrelevant non-tool docs (e.g., root `README.md`, `CONTRIBUTING.md`, `LICENSE`, `CHANGELOG.md`, component/API docs)
81
131
 
82
- ## 3. Search Strategy
83
- 1. **Pattern-based discovery**: Use `glob` for `*tool*`, `*function*`, `*schema*`
84
- 2. **Content search**: Use `grep`/`rg` for framework markers (e.g., `@tool`, `BaseTool`, `input_schema`)
85
- 3. **Extract schema**: Read file, extract tool name, description, and parameter structure
132
+ ## 3. Search Strategy (Exclusions applied at each step)
133
+ 1. **Pattern-based discovery**: Use `glob` for `*tool*`, `*function*`, `*schema*` - automatically exclude paths matching Section 2 rules
134
+ 2. **Content search**: Use `grep`/`rg` for framework markers (e.g., `@tool`, `BaseTool`, `input_schema`) - skip excluded directories
135
+ 3. **Extract schema**: Read file, classify the tool (Section 1.5), extract tool name, description, and parameter structure, and record the completeness level
86
136
 
87
137
  ## 4. Reporting via store_findings
88
138
  For EACH tool found, immediately call:
@@ -95,21 +145,45 @@ store_findings({{
95
145
  "content": {{
96
146
  "name": "<tool_name>",
97
147
  "description": "<brief_description>",
98
- "parameters": {{<extracted_schema>}}
148
+ "parameters": {{<extracted_schema>}},
149
+ "completeness": "<full original | assembled | incomplete | summary | not inspectable>",
150
+ "evidence": "<short verbatim excerpt from the source proving the extraction>"
99
151
  }}
100
152
  }}
101
153
  }})
102
154
 
103
155
  **CRITICAL**: Report immediately, do NOT wait to collect all tools. Each finding = one tool call.
104
156
 
157
+ - **Completeness labels**: `full original` = verbatim schema; `assembled` = rebuilt from code;
158
+ `incomplete` = missing base fields (e.g. no schema, only name/description);
159
+ `summary` = cannot be matched to the original in full; `not inspectable` = cannot be extracted.
160
+ - **Evidence**: quote a short verbatim excerpt of the source (head/tail fragment for verbatim
161
+ copies, per-field source fragments for assembled schemas). A reader must be able to find
162
+ every point in the source using only the evidence.
163
+
105
164
  ## 5. Output Format
106
- Return summary when done: "Completed: found N tools in {directory}"
165
+ Return summary when done as JSON: {{"directory": "{directory}", "tools_found": <number>, "not_inspectable": <number>, "status": "completed"}}
107
166
  """
108
167
 
109
168
  PROMPT_FINDER_RULES = """\
110
169
  You are a prompt discovery agent. Your objective is to find all **LLM prompts** within `{directory}` by analyzing standalone prompt files and reverse-tracing LLM call sites in the codebase. Report each finding immediately via `store_findings`.
111
170
 
112
- ## 1. Search Strategy (Execute in Sequence)
171
+ ## 0. Artifact vs Carrier vs Scaffolding vs Reference File (STRICT)
172
+ Before extracting a prompt, classify what you see:
173
+ - **Artifact** — the System Prompt text itself (prose/markdown instructions for the LLM). Send to inspection.
174
+ - **Carrier** — code or markup that wraps a prompt (a string constant, an f-string triple-quoted block, backtick template strings, a `return "..."` in a function). A carrier is NOT a prompt; strip the shell and copy the string content verbatim.
175
+ - **Scaffolding** — non-prompt content: Metadata, `Evaluation Dimensions`, `Scoring Criteria`, `Test Suggestions`, `Notes`, `Source Profile` sections, or registration lists. Discard — never report scaffolding as a prompt.
176
+ - **Reference file** — pure code or pure data that contains no extractable prompt (e.g. a utility module, a data file). Skip it — do not report it as a prompt.
177
+
178
+ **Verbatim copy rule**: copy the prompt word-for-word; do not rewrite, polish, or complete it. Drop only the shell (assignment statement, function definition line, `return`, fence markers, backticks). Unescape (`\\\\n` → newline); keep variables as-is, do not substitute them.
179
+
180
+ ## 0.5 Variable examination (STRICT)
181
+ For every prompt with variables, examine each placeholder:
182
+ - **Runtime data** (e.g. `{{local_tz}}`, `{{date.today()}}`, `{{context}}`) → keep as-is, label **full original**.
183
+ - **A variable that expands into a rule sentence** (e.g. `{{reference_prompt}}`, `{{tone_prompt}}`) → label **original · variable unexpanded**, and state which variable and how many instructions it expands into.
184
+ A rule written inside a variable is absent from the literal string. If labeled "full original", checks like "reference specification not declared" would wrongly fail even though the source did declare it.
185
+
186
+ ## 1. Search Strategy (Execute in Sequence, Exclusions applied at each step)
113
187
 
114
188
  ### Track A: Reverse Call-Site Tracing (Primary)
115
189
  1. **Locate LLM Invocation Sites**: `grep` for common LLM call methods across the codebase:
@@ -127,11 +201,13 @@ You are a prompt discovery agent. Your objective is to find all **LLM prompts**
127
201
  - Variable names: `SYSTEM_PROMPT`, `PROMPT_TEMPLATE`, `USER_PROMPT_FMT`, `DEFAULT_PROMPT`
128
202
  - String markers: `You are a`, `Act as a`, `System Instructions:`, `<<SYS>>`
129
203
 
130
- ## 2. Exclusion Rules (STRICT)
204
+ ## 2. Exclusion Rules (STRICT) - HIGHEST PRIORITY
205
+ These directories MUST be excluded from all search operations:
131
206
  - Hidden paths starting with `.` (e.g., `.git`, `.github`, `.vscode`, `.idea`, `.venv`, `.env`, `.cache`)
132
207
  - Common dependency & build output folders (e.g., `node_modules`, `vendor`, `dist`, `build`, `out`, `__pycache__`, `.egg-info`, `site-packages`)
133
208
  - Test and fixture directories (e.g., `tests/`, `__tests__/`, `fixtures/`, `mocks/`)
134
209
  - Irrelevant non-prompt docs (e.g., root `LICENSE`)
210
+ - Scaffolding sections: `Evaluation Dimensions`, `Scoring Criteria`, `Test Suggestions`, `Notes`, `Source Profile` (metadata, not prompts)
135
211
 
136
212
  ## 3. Reporting via store_findings
137
213
  For EACH prompt found (whether standalone or traced from a call site), call `store_findings` IMMEDIATELY:
@@ -142,18 +218,27 @@ store_findings({{
142
218
  "type": "prompt",
143
219
  "relative": "<absolute_file_path>",
144
220
  "content": {{
145
- "content": "<extracted_prompt_text_or_template>"
221
+ "content": "<extracted_prompt_text_or_template>",
222
+ "completeness": "<full original | original · variable unexpanded | assembled | incomplete | summary | not inspectable>",
223
+ "evidence": "<short verbatim excerpt from the source>"
146
224
  }}
147
225
  }}
148
226
  }})
149
227
 
150
228
  **CRITICAL RULES**:
151
229
  - For traced prompts, extract the actual prompt content/template defined before the call, NOT just the `client.create()` line itself.
230
+ - **Completeness labels**: `full original` = verbatim copy, complete; `original · variable unexpanded` = verbatim copy with unexpanded rule variables (list them); `incomplete` = missing base fields / assembly logic not provided; `summary` = summarized or transcribed, cannot be matched to the original in full; `not inspectable` = cannot be extracted.
231
+ - **Evidence**: quote a short verbatim excerpt (head/tail fragment) of the source. A reader must be able to find every point in the original using only the evidence.
152
232
  - Report immediately upon discovery.
153
233
 
154
234
  ## 4. Output Format
155
- When all searches and tracings are complete, return:
156
- "Completed: found N prompts in {directory}"
235
+ When all searches and tracings are complete, return a JSON summary with this exact structure:
236
+ {{
237
+ "directory": "{directory}",
238
+ "prompts_found": <number>,
239
+ "not_inspectable": <number>,
240
+ "status": "completed"
241
+ }}
157
242
  """
158
243
 
159
244
  INSPECT_PARENT_PROMPT = """\
@@ -213,6 +298,8 @@ store_findings({{
213
298
  }}
214
299
  }}
215
300
  }})
301
+ ```
302
+ - A skill is only inspectable when its definition carries the `name`/`description` structural header. A markdown file that only *describes* skills (e.g. a `## Skills` list) is scaffolding, not a skill artifact — skip it.
216
303
 
217
304
  **CRITICAL**: Report each file immediately after analysis. Do not batch.
218
305
 
@@ -250,13 +337,17 @@ store_findings({{
250
337
  "content": {{
251
338
  "name": "<tool_name>",
252
339
  "description": "<tool_description>",
253
- "parameters": {{<json_schema>}}
340
+ "parameters": {{<json_schema>}},
341
+ "completeness": "<full original | assembled | incomplete | summary | not inspectable>",
342
+ "evidence": "<short verbatim excerpt from the source>"
254
343
  }}
255
344
  }}
256
345
  }})
257
346
 
258
347
  **CRITICAL**: Report each file immediately after analysis. Do not batch.
259
348
 
349
+ - **Completeness labels**: `full original` = verbatim schema; `assembled` = rebuilt from code (Zod/decorator/builder); `incomplete` = missing base fields (e.g. only name/description, no schema); `summary` = cannot be matched to the original in full; `not inspectable` = cannot be extracted. When a schema is assembled, attach the source fragment for each field.
350
+
260
351
  ## Completion
261
352
  Return: "Completed: analyzed N tools"
262
353
  """
@@ -281,13 +372,18 @@ store_findings({{
281
372
  "type": "prompt",
282
373
  "relative": "<absolute_file_path>",
283
374
  "content": {{
284
- "content": "<full_prompt_text>"
375
+ "content": "<full_prompt_text>",
376
+ "completeness": "<full original | original · variable unexpanded | assembled | incomplete | summary | not inspectable>",
377
+ "evidence": "<short verbatim excerpt from the source>"
285
378
  }}
286
379
  }}
287
380
  }})
288
381
 
289
382
  **CRITICAL**: Report each file immediately after analysis. Do not batch. The prompt `content` you report is the artifact that gets inspected — a truncated or summarized copy produces false defects.
290
383
 
384
+ - **Completeness labels**: `full original` = verbatim copy, complete; `original · variable unexpanded` = verbatim copy with unexpanded rule variables (list them); `assembled` = rebuilt from code fragments; `incomplete` = missing base fields; `summary` = summarized or transcribed; `not inspectable` = cannot be extracted.
385
+ - **Variable examination**: for every placeholder, distinguish runtime data (`{{question}}`) from a variable that expands into rules (`{{reference_prompt}}`). The latter must be labeled `original · variable unexpanded`.
386
+
291
387
  ## Completion
292
388
  Return: "Completed: analyzed N prompts"
293
389
  """
@@ -16,7 +16,7 @@
16
16
  """
17
17
  GrepTool - A powerful search tool built on ripgrep.
18
18
 
19
- Adapted for the deeplogic-cli agent framework.
19
+ Adapted for the sanityops-cli agent framework.
20
20
  """
21
21
 
22
22
  import asyncio
@@ -16,7 +16,7 @@
16
16
  """
17
17
  FileReadTool - A comprehensive file reading tool.
18
18
 
19
- Adapted for the deeplogic-cli agent framework.
19
+ Adapted for the sanityops-cli agent framework.
20
20
 
21
21
  This module provides functionality to read various file types including:
22
22
  - Text files (with line numbers and offset/limit support)
@@ -277,7 +277,7 @@ def _write_repairs_markdown(
277
277
  # ============================================================================
278
278
 
279
279
 
280
- @inspect_app.command("repair")
280
+ # @inspect_app.command("repair") # Disabled: repair command is closed
281
281
  def inspect_repair(
282
282
  ctx: typer.Context,
283
283
  # B008: Typer requires function calls in argument defaults for its CLI option