dkatsiros

Talos — Productivity skill for Claude Code

Productivity community

File-driven task queue for delegating work to autonomous Claude Code engineering agents.

How to install Talos

This entry records only its repository, not the path inside it, so there is no exact command to give. Open dkatsiros/talos and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Talos does

File-driven task queue for delegating work to autonomous Claude Code engineering agents.

Alternatives in Productivity

Skill Context Detection — Auto-detect work context (Dev vs Knowledge) — use to tailor workflows based on current task type 2.8k ★
Build In Public — Build-in-Public — generate posts from your engineering work, anonymize, and schedule via Buffer 705 ★
Agentacct — See what your coding agents did and what it cost 665 ★

README

Talos

*Python package & CLI: `openclaw-claude-loop`*

A **file-driven task queue** that lets an orchestrator delegate work to per-project [Claude Code](https://claude.ai/code) sessions acting as autonomous engineering agents ("CTO", "builder", "QA", etc.).

Each delegated task gets a structured contract: the orchestrator enqueues a task with title, instructions, role, and constraints; the engine renders a prompt, spawns a non-interactive `claude -p` session in a detached `screen`, waits for `result.json`, validates it, and returns the parsed result. Task history lives under `.openclaw/claude-loop/` inside the project it targets — completely separate from your orchestrator's state.

Install

pip install openclaw-claude-loop
# or for development:
pip install -e /path/to/this-repo

**Requires:** Python ≥ 3.10, `claude` CLI on `PATH` (Claude Code), `screen`.

Quick start (Python API)

from openclaw_claude_loop import ProjectLoop

loop = ProjectLoop("/path/to/your/project")

# Bootstrap the queue structure once per project
if not loop.is_bootstrapped:
    loop.bootstrap()

# Delegate a task — blocks until result.json arrives
result = loop.run(
    title="Add /healthcheck endpoint",
    instructions=[
        "Add GET /healthcheck to app/main.py returning {build_sha, timestamp}",
        "Add pytest in tests/test_healthcheck.py",
        "Run: pytest tests/test_healthcheck.py -v",
    ],
    role="cto",      # 'cto' | 'builder' | 'qa' | 'backend' | 'frontend' | ...
    timeout=900,
)
print(result["state"], result["summary"])

CLI

# Bootstrap a project
talos --project-root /path/to/project bootstrap

# Enqueue a task
talos --project-root /path/to/project enqueue "Add /healthcheck" \
    --role cto \
    --instruction "Add GET /healthcheck to app/main.py returning {build_sha, timestamp}" \
    --instruction "Write pytest in tests/test_healthcheck.py and run it"

# Render the prompt (parks the task ready for hand-