Caid Multi Agent
v1.2.0Coordinate multiple sub-agents to collaboratively complete long-horizon software engineering tasks using the CAID (Centralized Asynchronous Isolated Delegati...
⭐ 0· 41·0 current·0 all-time
bySimon@simoncatbot
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description (CAID multi-agent coordination) match the instructions: dependency analysis, physical git worktrees, delegation, test-gated integration, and spawning engineer subagents. All requested operations (git worktree, venv, pytest, spawn subagents) are expected for this capability.
Instruction Scope
SKILL.md gives detailed runtime steps that stay within the domain (repo prep, create worktrees, delegate tasks, run tests, merge). It also includes sample code to call a sessions_spawn API (platform subagent spawn). The instructions assume the agent can run shell commands, access the repository, and call the platform's session API — these are expected but the document does not explicitly list the external privileges (git remote access, platform session API access) that will be used.
Install Mechanism
Instruction-only skill with no install spec or external downloads. It runs common tooling (python venv, pip, pytest, git) but does not install code from arbitrary URLs — low install risk.
Credentials
The skill declares no required environment variables or credentials, which is reasonable for an instruction-only coordination skill. However, the instructions call git push/pull (remote repository operations) and call sessions_spawn (platform API) — both implicitly require repository credentials and platform runtime permissions. The skill doesn't request or document these, so operators should ensure the environment already has appropriate git auth and platform session permissions before use.
Persistence & Privilege
always:false and no special install behavior. The skill instructs spawning subagents (normal for multi-agent coordination) but does not request permanent system presence or attempt to modify other skills' configs.
Assessment
This skill is internally consistent with its goal of coordinating multiple engineer agents. Before installing or using it, ensure: (1) the agent runtime has access to the target repository (local workspace and any git remote credentials needed for git push/pull), (2) the execution environment can create worktrees and virtual environments and run tests (python, pytest, required libs), and (3) the platform account used by the agent has permission to spawn subagents via the sessions_spawn-like API. If you don't want automatic remote pushes, run merges locally or mirror a safe remote. Limit the number of concurrent engineers according to the guidance (2–4 typical; 8 absolute max). If you need clarification about what platform tokens or git credentials will be used, ask the skill author or your platform admin before enabling it.Like a lobster shell, security has layers — review code before you run it.
asyncvk97a9mye5n5628524nmrsaep99840ekkcoordinationvk97a9mye5n5628524nmrsaep99840ekklatestvk97f3z35j9wphn0xzm2wegvg91841wnkmulti-agentvk97a9mye5n5628524nmrsaep99840ekk
License
MIT-0
Free to use, modify, and redistribute. No attribution required.