bmad-tech-spec
Pass
Audited by Gen Agent Trust Hub on Jun 20, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: Analysis of the skill instructions and the provided technical specification template revealed no evidence of malicious behavior, obfuscation, or high-risk command execution. The skill is scoped correctly for its documentation and planning purpose.
- [PROMPT_INJECTION]: The skill manages untrusted input which represents a surface for indirect prompt injection.
- Ingestion points: The workflow incorporates data from interactive user conversations and local project files such as
bmad-output/project-context.mdandbmad-output/tech-spec.mdinto the generated documents. - Boundary markers: The
tech-spec.template.mduses standard placeholders without explicit delimiters or guardrail instructions to isolate untrusted input from the document structure. - Capability inventory: The skill is configured with tools for file manipulation (
Read,Write,Edit,Glob,Grep) and information gathering (WebSearch,WebFetch). - Sanitization: While input is not programmatically sanitized, the skill implements a 'Validation Checklist' requiring the agent to confirm that 'No content instructs a dev agent to run tests, lint, build, or deploy', which acts as a deterrent against malicious instructions being persisted in the planning output.
Audit Metadata