Audit History
self-improving-agent - 8 audits
Version comparison
Capability and finding changes across audited versions, newest first.
| Version | Date | Result | Review items | Change vs previous |
|---|---|---|---|---|
| v8 Latest | Jul 23, 2026, 12:11 PM | 1 confirmed | 0 | No capability change |
| v7 | Jul 17, 2026, 11:06 AM | 1 confirmed | 0 | No capability change |
| v6 | Jul 17, 2026, 11:06 AM | 1 confirmed | 0 | No capability change |
| v5 | Jul 17, 2026, 11:06 AM | 1 confirmed | 0 | No capability change |
| v4 | Jul 6, 2026, 05:31 AM | 3 confirmed | 10 | No capability change |
| v3 | Jul 6, 2026, 05:31 AM | 3 confirmed | 10 | No capability change |
| v2 | Jun 28, 2026, 09:47 PM | 2 confirmed | 2 | External commandsNetwork accessFilesystem access |
| v1 | Mar 17, 2026, 08:23 AM | No confirmed findings | 0 | Baseline |
Jul 23, 2026, 12:11 PM
Static findings are predominantly false positives caused by Markdown code fences, inline backticks, research links, and documentation of local paths. The bundled hook scripts only log event metadata and output lengths; they do not perform network requests, command interpolation, or memory writes. The documented automatic memory workflow still needs privacy controls because it can retain context from many skill events.
Confirmed security concerns (1)
Risk Factors
⚙️ External commands (48)
🌐 Network access (12)
📁 Filesystem access (4)
Jul 17, 2026, 11:06 AM
Static findings are predominantly false positives caused by Markdown code fences, inline backticks, research links, and documentation of local paths. The bundled hook scripts only log event metadata and output lengths; they do not perform network requests, command interpolation, or memory writes. The documented automatic memory workflow still needs privacy controls because it can retain context from many skill events.
Confirmed security concerns (1)
Risk Factors
⚙️ External commands (48)
🌐 Network access (12)
📁 Filesystem access (4)
Jul 17, 2026, 11:06 AM
Static findings are predominantly false positives caused by Markdown code fences, inline backticks, research links, and documentation of local paths. The bundled hook scripts only log event metadata and output lengths; they do not perform network requests, command interpolation, or memory writes. The documented automatic memory workflow still needs privacy controls because it can retain context from many skill events.
Confirmed security concerns (1)
Risk Factors
⚙️ External commands (48)
🌐 Network access (12)
📁 Filesystem access (4)
Jul 17, 2026, 11:06 AM
Static findings are predominantly false positives caused by Markdown code fences, inline backticks, research links, and documentation of local paths. The bundled hook scripts only log event metadata and output lengths; they do not perform network requests, command interpolation, or memory writes. The documented automatic memory workflow still needs privacy controls because it can retain context from many skill events.
Confirmed security concerns (1)
Risk Factors
⚙️ External commands (48)
🌐 Network access (12)
📁 Filesystem access (4)
Jul 6, 2026, 05:31 AM
Most static external command and network alerts are false positives from Markdown examples, diagrams, and research links. Confirmed risks remain because the community skill uses hidden Claude configuration paths, symlink installation, optional Bash hooks, and broad self-modifying workflows that can persist sensitive session context.
Confirmed security concerns (3)
Capability review items (10)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
Risk Factors
⚙️ External commands (37)
🌐 Network access (10)
Jul 6, 2026, 05:31 AM
Most static external command and network alerts are false positives from Markdown examples, diagrams, and research links. Confirmed risks remain because the community skill uses hidden Claude configuration paths, symlink installation, optional Bash hooks, and broad self-modifying workflows that can persist sensitive session context.
Confirmed security concerns (3)
Capability review items (10)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
Risk Factors
⚙️ External commands (37)
🌐 Network access (10)
Jun 28, 2026, 09:47 PM
Static analysis produced many alerts, but most blocker-level items are false positives from markdown examples, diagrams, and research links. The confirmed risk is high because this community skill requests broad file-editing and Bash capabilities, teaches self-modification of skills, and documents hooks that can log tool input and command output.
Confirmed security concerns (2)
Capability review items (2)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
Static false positives ignored (2)
These static matches were dismissed by semantic review or matched schema-only tokens, so they are shown for transparency but do not drive the quality score.
Risk Factors
⚙️ External commands (3)
🌐 Network access (3)
📁 Filesystem access (3)
Detected Patterns
Mar 17, 2026, 08:23 AM
This is a legitimate Claude Code skill for self-improvement. The static scanner flagged 121 potential issues, but after semantic evaluation, all are false positives. External command detections are markdown backticks in documentation. Network detections are academic reference URLs. Filesystem detections are standard Claude Code config paths. No actual security risks present.