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.
62
56%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Critical
Do not install without reviewing
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/task-orchestrator/SKILL.mdQuality
Discovery
85%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
The description is technically strong with specific capabilities and an explicit 'Use for...' clause that clearly delineates when to apply the skill. Its main weakness is that trigger terms lean heavily on technical jargon rather than natural user language—users needing this skill might phrase their requests differently (e.g., 'work on multiple tasks simultaneously' or 'split this into parallel subtasks'). Overall it's a solid description that would benefit from more natural-language trigger variations.
Suggestions
Add more natural-language trigger terms users might actually say, such as 'split work into parallel tasks', 'work on multiple things at once', 'batch process issues', or 'run subtasks simultaneously'.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: 'dependency analysis', 'parallel tmux/Codex execution', 'self-healing heartbeat monitoring', and 'multi-agent task orchestration'. These are concrete, identifiable capabilities. | 3 / 3 |
Completeness | Clearly answers both what ('autonomous multi-agent task orchestration with dependency analysis, parallel tmux/Codex execution, and self-healing heartbeat monitoring') and when ('Use for large projects with multiple issues/tasks that need coordinated parallel execution'). | 3 / 3 |
Trigger Term Quality | Includes some relevant terms like 'parallel execution', 'multiple issues/tasks', and 'large projects', but uses more technical jargon ('tmux', 'heartbeat monitoring', 'orchestration') than natural user language. Users are more likely to say 'work on multiple tasks at once' or 'parallelize work' than 'multi-agent task orchestration'. | 2 / 3 |
Distinctiveness Conflict Risk | Highly distinctive with a clear niche: multi-agent parallel orchestration with tmux/Codex and heartbeat monitoring. This is unlikely to conflict with other skills due to its very specific combination of capabilities. | 3 / 3 |
Total | 11 / 12 Passed |
Implementation
27%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill attempts to document a complex multi-agent orchestration system but suffers from significant verbosity, inlining too much detail that should be split into referenced files. While it provides many concrete bash snippets, several are incomplete templates or use invalid syntax, reducing actionability. The workflow structure is reasonable but lacks the bounded retry validation and proper checkpoints that the skill's own safety boundaries demand.
Suggestions
Extract the manifest JSON schema, monitoring script, and example orchestration into separate referenced files (e.g., MANIFEST_SCHEMA.md, MONITOR.sh, EXAMPLES.md) to dramatically reduce the main file size and improve progressive disclosure.
Remove explanations of concepts Claude already knows (what worktrees are, what tmux sessions do, basic dependency concepts) and trim the Core Concepts section to just the dependency rules table.
Fix the pseudo-bash in Step 3 (for-loops with undefined syntax) to be either valid bash or clearly marked as pseudocode patterns, and add explicit bounded retry logic to match the safety boundary about not retrying indefinitely.
Add explicit validation checkpoints: verify worktree creation succeeded, verify Codex process started in tmux, verify push succeeded before creating PR.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is extremely verbose at ~300+ lines. It explains concepts Claude already knows (what git worktrees are, what tmux sessions do, basic JSON structure), includes a full manifest JSON schema that could be referenced externally, and repeats similar bash patterns multiple times. The 'Core Concepts' section explains dependency rules that are self-evident. Time-sensitive details like specific dates and model names (gpt-5.2-codex, 2026-01-17) add noise. | 1 / 3 |
Actionability | The skill provides many concrete bash commands and code snippets, but most are templates with placeholders (OWNER/REPO, $task_id) rather than fully executable code. Key steps like 'Analyze for dependencies (files mentioned, explicit deps)' are left as comments/pseudocode. The issue analysis step is vague ('Group by files mentioned in issue body'). The for-loop syntax in Step 3 is not valid bash. | 2 / 3 |
Workflow Clarity | The 6-step workflow is clearly sequenced with phase gates, and the monitoring section includes self-healing actions. However, validation checkpoints are weak — there's no explicit verification that worktrees were created correctly, no check that Codex actually started, and the self-healing retry has no bounded retry count despite the safety boundary mentioning it. The heartbeat cron syntax references a non-standard 'cron action:add' command without explanation. | 2 / 3 |
Progressive Disclosure | Everything is in one monolithic file with no references to external files for detailed content. The manifest schema, the full monitoring script, the example JSON, and lessons learned are all inline. The 'Integration with Other Skills' section mentions other skills but doesn't offload any content. The lessons learned section and tips could easily be separate files, and the monitoring script should be a referenced file. | 1 / 3 |
Total | 6 / 12 Passed |
Validation
72%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 8 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
metadata_version | 'metadata.version' is missing | Warning |
metadata_field | 'metadata' should map string keys to string values | Warning |
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 8 / 11 Passed | |
- Repository
- jdrhyne/agent-skills
- Commit
6768672
- Reviewed
Table of Contents
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.