sdlc-kit
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from multiple sources including feature requests, bug reports, project source code, and artifacts which are interpolated into prompts for tasks like intent analysis, specification drafting, and build execution.
- Ingestion points: The agent reads project-specific files such as intent.md, spec.md, plan.md, evidence.md, harvest.md, and memory files like POLICY.md, INDEX.md, and DOMAIN.md, as well as the underlying project source code during discovery and build stages.
- Boundary markers: While artifacts are formatted as Markdown, the skill instructions do not specify explicit delimiters or instructions to ignore potentially malicious embedded content within these files.
- Capability inventory: The skill possesses the capability to execute shell commands (as defined in .sdlc/config.md), perform git operations (commit, push, fetch), and spawn subagents for verification and adversarial review.
- Sanitization: There is no evidence of content sanitization or validation to filter out potential prompt injection attacks from the ingested text before processing.
Audit Metadata