tlc-implement
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes standard command-line tools for development and testing, including
npm,pytest,jest,git, andrg. These are used to run test suites, manage worktrees for fault injection, and search the codebase. The skill instructions include a 'Blast Radius' rule requiring explicit user go-ahead for operations likegit pushor production deployments. - [INDIRECT_PROMPT_INJECTION]: The skill acts on untrusted external data by extracting implementation requirements from sources like tickets, PRDs, and RFCs. This creates a surface for instructions hidden in documentation to influence the agent.
- Ingestion points: The 'Extract' phase in
SKILL.mdrequires reading external documentation, PRDs, and RFCs. - Boundary markers: None present; the skill does not explicitly instruct the agent to use delimiters or ignore instructions inside the ingested data.
- Capability inventory: The skill allows filesystem writes to a
.checks/directory, execution of shell-based test runners, and the ability to dispatch sub-agents. - Sanitization: No sanitization or validation of the external text content is specified.
Audit Metadata