Skills
Navigate
skills/full-stack-skills/chart-skills/openspec-apply-change/Gen Agent Trust Hub

openspec-apply-change

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the openspec CLI tool to perform repository management and task retrieval, which is the intended purpose of the skill.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external inputs which represent a potential attack surface for indirect prompt injection. 1. Ingestion points: JSON fields context and operationGuidance from CLI output and the contents of files listed in contextFiles. 2. Boundary markers: The skill instructs the agent to treat these inputs as advisory behavior contracts and to prioritize CLI-enforced logic and safety states. 3. Capability inventory: The skill uses Bash(openspec:*) and file system read/write operations. 4. Sanitization: The agent is instructed to report conflicts and avoid copying guidance into implementation artifacts.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 06:56 AM