Skills
skills/github/gh-aw/pr-to-go-linter/Gen Agent Trust Hub

pr-to-go-linter

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill ingests untrusted data from GitHub pull request diffs and metadata to generate new code. An attacker could craft a malicious PR designed to influence the code generation or the agent's behavior.
  • Ingestion points: Workflow steps 1 and 2 in SKILL.md read PR metadata and diffs.
  • Boundary markers: None defined to isolate external data from the implementation logic.
  • Capability inventory: The skill performs go test, go build, and make commands as documented in the Workflow section of SKILL.md.
  • Sanitization: No validation or sanitization of the PR content is described before it is used to generate code that the agent subsequently executes.
  • [COMMAND_EXECUTION]: The skill executes shell commands to validate the generated linter, specifically go test ./pkg/linters/<name>/..., go build ./cmd/linters, and make golint-custom.
  • [REMOTE_CODE_EXECUTION]: Because the skill executes go test and go build on code derived from external PR patterns, it could lead to the execution of malicious code (e.g., via Go init() functions or test logic) if the source PR content is adversarial.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 05:54 AM
Security Audit — agent-trust-hub — pr-to-go-linter