apply-grant
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes multiple bash blocks for telemetry and invokes a bundled shell script to export session logs. Evidence: Bash blocks in SKILL.md manage telemetry state in ~/.superstack/ and the workflow executes bash ~/.claude/skills/apply-grant/export-session.sh to copy conversation logs.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted project context to generate a grant application draft, presenting a vulnerability surface.
- Ingestion points: idea-context.md, build-context.md, README.md, package.json, Cargo.toml, Anchor.toml, and git log output.
- Boundary markers: Absent. There are no delimiters or instructions to the model to ignore potential commands embedded in the project files.
- Capability inventory: Shell execution (bash, curl, git), file system read/write, and network access for telemetry.
- Sanitization: Absent. No validation or filtering is performed on data read from external files before processing.
Audit Metadata