PRD for AI Agents
PassAudited by ClawScan on May 14, 2026.
Overview
The visible artifacts describe an instruction-only PRD-writing workflow with no code, install steps, credentials, or network behavior; users should review any generated agent instruction files before reusing them.
This skill appears safe to install as an instruction-only PRD generator. Share only project details or documents you are comfortable turning into a saved PRD, and review generated companion files like CLAUDE.md or AGENTS.md before letting other coding agents follow them.
Findings (1)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Generated PRDs and companion agent files may influence later coding-agent behavior, so inaccurate assumptions or unreviewed source material could affect future work.
These files are intended to persist and may be read as instructions or context by future coding agents. This is purpose-aligned and user-directed, but it can carry forward assumptions or errors.
If the user asks for companion files, also generate: - `CLAUDE.md` or `AGENTS.md` with project conventions from the PRD. - `TASKS.md` with the build order expanded into discrete tasks.
Review the final PRD and any CLAUDE.md, AGENTS.md, TASKS.md, or PLANNING.md files before using them with a coding agent, especially assumptions marked by the skill.