branch-context

Pass

Audited by Gen Agent Trust Hub on Oct 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a utility for managing local project context and does not exhibit any malicious patterns. It operates strictly within the local project environment.
  • [COMMAND_EXECUTION]: The skill provides helper scripts to manage its state files. These scripts (status.sh, append-handoff.sh, append-pr-decision.sh) use standard bash utilities to read, write, and format markdown content. They include appropriate sanitization, such as slugging filenames and escaping JSON output, to ensure robust operation.
  • [INDIRECT_PROMPT_INJECTION]: The skill manages a surface for indirect instructions by reading externalized context.
  • Ingestion points: Reads issue-brief.md, pr-decisions.md, handoffs-index.md, and session handoff files as specified in SKILL.md.
  • Boundary markers: Instructions explicitly mandate verbatim quoting of constraints and promises to maintain instruction modality across sessions.
  • Capability inventory: Scripts perform file writes and template instantiation using cat, cp, and printf. No network access or elevated privileges are requested.
  • Sanitization: append-handoff.sh implements filename slugging for generated handoffs, and append-pr-decision.sh validates that source references are URLs.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 6, 2026, 11:18 PM