Skills
Navigate
skills/kdy1/kdy1-scripts/slop-fix-batch/Gen Agent Trust Hub

slop-fix-batch

Pass

Audited by Gen Agent Trust Hub on Oct 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from GitHub issue bodies and comments.
  • Ingestion points: scripts/collect-issues.mjs and scripts/github-request.mjs fetch issue data and metadata from the GitHub API.
  • Boundary markers: The instructions in references/worker-prompts.md and references/construction.md explicitly instruct agents to treat issue content as evidence only and not as authority to execute instructions or expand the batch scope.
  • Capability inventory: The skill uses spawnSync in scripts/batch-stack.mjs to execute git and gh commands. These commands are executed using argument arrays, which prevents shell injection.
  • Sanitization: The skill performs strict validation of GitHub PR states, SHAs, and repository identities in scripts/verify-publication.mjs and scripts/batch-ledger.mjs before recording progress or allowing transitions. This limits the impact of potentially malicious data in the GitHub environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 15, 2026, 01:33 AM