tracer-convince-me
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DYNAMIC_EXECUTION]: The skill instructs the agent to 'supplement [tests] with a focused test, script, fixture, manual interaction, screenshot, or disposable harness' to verify claims. This involves runtime generation and execution of scripts, which is a core function of the skill's verification purpose. The instructions include a safety directive to 'Get approval before destructive actions, external side effects, production access, or spending money.'
- [INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection as it is designed to ingest and act upon external, potentially untrusted data such as 'the original request, available spec or ticket, [and] acceptance criteria' (SKILL.md).
- Ingestion points: The skill reads requests, specs, and tickets from the local environment or project context (SKILL.md).
- Boundary markers: The instructions do not define specific delimiters for untrusted data, though they do instruct the agent to translate expectations into observable claims before execution.
- Capability inventory: The skill has the capability to 'Run every proving command or interaction' and create scripts (SKILL.md).
- Sanitization: There are no explicit instructions for sanitizing or escaping the content of the ingested tickets or specs.
Audit Metadata