platonic-code-review
Pass
Audited by Gen Agent Trust Hub on Mar 8, 2026
Risk Level: SAFENO_CODEPROMPT_INJECTION
Full Analysis
- [SAFE]: Analysis of the skill's logic and templates confirmed the absence of malicious instructions, obfuscation, or attempts to bypass security filters.\n- [NO_CODE]: This skill is entirely comprised of Markdown files and reporting templates; it does not distribute any executable code, binaries, or dependency manifests.\n- [PROMPT_INJECTION]: The skill interacts with untrusted external files (specifications and source code), which introduces a potential surface for indirect prompt injection.\n- Ingestion points: Specification documents (RFCs, requirements) and source code are read during the review process as specified in SKILL.md and references/REFERENCE.md.\n- Boundary markers: The skill does not define explicit delimiters to isolate the content of these files from the agent's core instruction set.\n- Capability inventory: The agent has read access to the project files and conditional write access to the filesystem, which is restricted by a requirement for explicit user confirmation.\n- Sanitization: No sanitization or filtering logic is provided for the data extracted from the processed documents.
Audit Metadata