To orchestrate parallel coding-agent farms (Claude, Codex, Copilot, Gemini, etc.) on isolated git worktrees.
71%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Medium
Suggest reviewing before use
Fix and improve this skill with Tessl
tessl review fix ./instructions/r3/core/skills/coding-agents-farm/SKILL.mdThe skill prompts the agent to compromise the security or integrity of the user’s machine by modifying system-level services or configurations, such as obtaining elevated privileges, altering startup scripts, or changing system-wide settings.
The skill instructs the agent to disable permission prompts and sandboxes across several CLIs (bypassPermissions, --yolo, danger-full-access, CODEX_UNSAFE_ALLOW_NO_SANDBOX=1) and to write auto-approve settings into user-level config files like ~/.gemini/settings.json and ~/.config/goose/profiles.yaml, which is a security bypass that modifies machine state, though a user-confirmation gate partially mitigates it.
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
The required workflow never fetches outsider content itself; the only outsider-reachable path is that spawned Claude runs, with auto-approved WebFetch and MCP access (line 45), may read arbitrary web pages, and their terminal output is later read by the Monitor step (lines 143-151).
griddynamics/rosetta
ca35acf
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.