conversation-context-export
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands to interact with the local git environment and the GitHub API. It uses the
gitbinary to retrieve branch and worktree information, and thegh(GitHub CLI) tool to manage Pull Request comments. - [INDIRECT_PROMPT_INJECTION]: The skill serves as a bridge for conversation context across different agent sessions, creating a surface where instructions or data from one session could influence a downstream agent.
- Ingestion points: The skill ingests the current conversation history and reads existing context files from the
.dev/contexts/directory (Step 2 in SKILL.md). - Boundary markers: The skill uses a structured Markdown template (TEMPLATE.md) to organize the exported data, providing some structural delimitation, though it lacks explicit 'ignore' instructions for consuming agents.
- Capability inventory: The skill has the capability to write files to the local disk via the
Writetool and perform network operations through the GitHub CLI to create or update PR comments. - Sanitization: Branch names are sanitized by replacing special characters (
/ \ : * ? " < > |) with hyphens before being used in file paths, which helps prevent directory traversal or file naming issues.
Audit Metadata