briefing-followup

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill interacts with well-known Google API endpoints (gmail.googleapis.com and googleapis.com) to manage email drafts and calendar events. These network operations are performed via curl using standard OAuth access tokens provided by the environment.\n- [COMMAND_EXECUTION]: The skill orchestrates several local CLI tools to perform its tasks. It uses git for document versioning and commit management, linearis for interacting with Linear project management, and the claude CLI for background task dispatching. It also supports interactive editing of proposals via the system's $EDITOR variable. All commands are constructed using jq to ensure safe parameter handling.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from briefing markdown files.\n
  • Ingestion points: Briefing files at thoughts/briefings/YYYY-MM-DD.md and compound ADR proposals at thoughts/shared/compound/pending/*.md.\n
  • Boundary markers: The skill uses strict YAML frontmatter parsing and JSON schema validation (briefing-frontmatter.schema.json) to delimit data.\n
  • Capability inventory: File system writes, git commits/pushes, network API calls, and subprocess execution via CLI tools.\n
  • Sanitization: The skill employs jq for shell command interpolation and python3 for structured YAML/JSON processing, reducing the risk of data being misinterpreted as commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 09:29 PM
Security Audit — agent-trust-hub — briefing-followup