hrt-backlog-reconcile
Pass
Audited by Gen Agent Trust Hub on Aug 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from project specification files to suggest modifications to a backlog file.
- Ingestion points: Reads content from
openspec/specs/**/*.mdandopenspec/BACKLOG.mdwithin the SKILL.md instructions. - Boundary markers: Incorporates an interactive confirmation step (Step 3) where the agent must ask the user to confirm every proposed deletion.
- Capability inventory: The skill is limited to reading and writing specific markdown files in the
openspec/directory. No network or subprocess capabilities are utilized. - Sanitization: Relies on manual user review of each proposed deletion, providing a human-in-the-loop check against unintended actions derived from external file content.
- [NO_CODE]: The skill consists entirely of markdown instructions and does not include any executable scripts, libraries, or binary files.
Audit Metadata