Claude Skill MartBrowse skillsQuick linesLearn by videoTerminal guideWhat is a Skill?
 
← Back to list

Agent Native (Parallel Worker Dispatch)

An orchestration-discipline skill for dispatching independent, non-overlapping tasks to parallel workers or named role sessions while collecting observed runtime evidence.

AutomationAdvanced★ 447⑂ 41AI score 6/10Last updated: Oct 3, 2026

What it does

  • Splits authorized work into explicit roles — Orchestrator / Implementer / Validator / Scribe — plus cheap bounded Reader/Writer delegations.
  • Validates every packet before the first launch: selected executor, packet identity, resolved symlinks, normalized paths, and case-insensitive write-scope comparison to catch hidden collisions.
  • Dispatches each validated packet exactly once and keeps packet identity attached to the result (candidate, evidence, or executor error).
  • Requires observed session/context identity rather than requested IDs, and forbids promoting provider retries, reconnects, or idle states into plans or verdicts.
  • Names concrete failure modes: "prompt-send optimism" (delivery ≠ engagement) and restarting a stalled worker before capturing its state.

Who it's for

  • Teams already on the AgentOps workflow, or anyone building pipelines that fan out to multiple coding agents.
  • Engineers who keep hitting "the prompt was delivered but nothing is happening" failures.
  • Anyone who needs hard guarantees that concurrent writers cannot touch the same files.

Example uses

  1. Build three packets editing three separate modules, prove write scopes are disjoint, then launch them in parallel.
  2. On a suspected stall, capture observable worker state first so the stall evidence survives instead of being destroyed by a restart.
  3. Keep a large file out of the parent context by delegating a ≤350-line slice to a Reader role and receiving only line-referenced bullets.

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/boshu2/agentops/HEAD/skills/agent-native/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the skills/agent-native folder from the GitHub repo boshu2/agentops into my ~/.claude/skills/boshu2-agent-native-2/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/boshu2/agentops.git && mkdir -p ~/.claude/skills && cp -r agentops/skills/agent-native ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

  1. Open a terminal.
  2. Clone the repository: git clone https://github.com/boshu2/agentops.git
  3. Create the skills directory: mkdir -p ~/.claude/skills
  4. Copy the skill: cp -r agentops/skills/agent-native ~/.claude/skills/
  5. Confirm the references/ documents (model-dispatch.md, context-budget-delegation.md, session-associations.md, etc.) came along with it.
  6. Note that scripts/swarm/dispatch_once.py is a developer reference requiring an AgentOps source checkout and is not bundled with standalone skills — keep the cloned repo if you need it.
  7. Restart Claude Code and try a request like "fan this out to parallel workers with disjoint write scopes" to verify the skill loads.
View source on GitHub ↗License: Apache-2.0