tailrocks-contribute-prepare
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill interacts with external repository content and proposal files, which introduces a potential surface for indirect prompt injection attacks.
- Ingestion points: The agent processes fork clones and contribution proposals as specified in
SKILL.md. - Boundary markers: The skill includes comprehensive mitigation policies in
references/runtime-trust.mdandreferences/contribution-handoff.md, which explicitly instruct the agent to treat external content as untrusted data and ignore any embedded instructions. - Capability inventory: Capabilities are restricted to local file operations and Git commits; the skill explicitly prohibits network access, credential use, and upstream mutations.
- Sanitization: The accompanying script
scripts/contribute-prepare.tsperforms rigorous input validation, checking for absolute paths, resolving real paths, and explicitly rejecting symbolic links to prevent path traversal or filesystem attacks. - [EXTERNAL_DOWNLOADS]: The skill script references a core library located outside the skill's immediate directory.
- Evidence:
scripts/contribute-prepare.tsimportscontributionStageClifrom../../../scripts/contribution-stage-core. This is a vendor-provided internal resource consistent with the skill's infrastructure and does not represent an untrusted external download.
Audit Metadata