task-orchestrator
Autonomous multi-agent task orchestration with dependency analysis, parallel tmux/Codex execution, and self-healing heartbeat monitoring. Use for large projects with multiple issues/tasks that need coordinated parallel execution.
71
88%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
View guide
Critical
Do not install without reviewing
Security
2 findings: 1 critical severity, 1 high severity. Installing this skill is not recommended: please review these findings carefully if you do intend to do so.
E006: Malicious code pattern detected in skill scripts.
What this means
Detected high-risk code patterns in the skill content — including its prompts, tool definitions, and resources — such as data exfiltration, backdoors, remote code execution, credential theft, system compromise, supply chain attacks, and obfuscation techniques.
Why it was flagged
This skill enables fully autonomous code modification and deployment (auto-responding to prompts, restarting agents, pushing branches, creating PRs, and running via cron) and includes guidance to bypass sandbox/network limits — patterns that strongly enable supply-chain/backdoor insertion and unauthorized remote code changes even though it does not explicitly show data exfiltration or an external C2 endpoint.
W007: Insecure credential handling detected in skill instructions.
What this means
The skill handles credentials insecurely by requiring the agent to include secret values verbatim in its generated output. This exposes credentials in the agent’s context and conversation history, creating a risk of data exfiltration.
Why it was flagged
The orchestrator captures tmux/log output into files and explicitly injects that content verbatim into prompts/commands sent to Codex (e.g., via $(cat error.log | tail -20)), which would expose any secrets in logs to the LLM and downstream outputs.
Low
Low-risk findings.
2 low severity findings. Worth noting, but not necessarily harmful.
W011: Third-party content exposure detected (indirect prompt injection risk).
What this means
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.
Why it was flagged
This skill explicitly fetches open GitHub issues (see "gh issue list --repo OWNER/REPO --state open --json number,title,body > issues.json" and "/tmp/issues.json" in the Workflow and "Analyze GitHub Issues for Dependencies" sections) and uses issue bodies to group tasks and build prompts that are sent to Codex, so untrusted, user-generated content from public issues directly influences agent decisions and tool actions.
W012: Unverifiable external dependency detected (runtime URL that controls agent).
What this means
The skill fetches instructions or code from an external URL at runtime, and the fetched content directly controls the agent’s prompts or executes code. This dynamic dependency allows the external source to modify the agent’s behavior without any changes to the skill itself.
Why it was flagged
The skill performs a runtime git clone of the target repository (git clone https://github.com/OWNER/REPO.git) and then runs Codex and test/command workflows inside the cloned worktrees, meaning fetched remote code from that URL is pulled at runtime and can be executed by the agent.
- Repository
- jdrhyne/agent-skills
- Commit
a4d31ad
- Audited
- Security analysis
Is this your skill?
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.