dispatch

A dispatch process for sending one approved implementation task to a coding worker in an isolated Git worktree. A worktree is a separate working copy of a repository.

In plain words
What is it for?
Use it to send an authorized task to a worker, or coordinate several tasks across branches in a single repository.
Why use it?
It keeps worker changes separated and uses built-in agent isolation instead of maintaining a custom queue for each task.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/lukasrepublic/agentic-foundry/dispatch
Any agent
npx skills add lukasrepublic/agentic-foundry --skill dispatch
Clone the repo
git clone --depth 1 https://github.com/lukasrepublic/agentic-foundry

Made for: Claude Code, Codex.

Per session 115 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 5,686 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00115 $0.05686
Opus 5 $0.00057 $0.02843
Sonnet 5 $0.00023 $0.01137
Haiku 4.5 $0.00012 $0.00569

Measured 2d ago against content hash 3d1212957f54, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

dispatch scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/dispatch/SKILL.md · 304 lines

How it starts

The opening of the file, as written. The whole thing — 304 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/foundry:dispatch

The lean dispatch path — a native-primitive REPLACEMENT for the bespoke dispatch stack. Worker spawn = the native Agent tool (isolation: "worktree", structured return); multi-atom fan-out = the native Workflow tool (workflows/release-wave.js). For a single-repo adopter this REPLACES the bespoke dispatch-agent + .claude/dispatch-queue/ machinery (queue manifests, flock, wt claim, WorktreeCreate prereqs, result.json polling, fan-out hygiene) — native primitives subsume it.

Single-repo only. Native isolation:worktree worktrees the session repo (the workspace). It CANNOT reach a separate gitignored product clone, so a multi-repo adopter (workspace ⟷ product-repo, e.g. acme-app, or self-hosting agentic-foundry/) still needs a re-extracted minimum: an explicit target_repo (a frozen+authorized contract field), a per-spawn dispatch manifest, the WorktreeCreate hook (foundry-worktree-create.sh) that redirects the worktree to the named product repo via foundry-wt, and worktree-time binding of the worker to that named repo (so its PR can only land there). See Multi-repo dispatch below. This is NOT the heavyweight bespoke queue — it is the one fact native isolation cannot provide (which repo to worktree).

Native default vs. the deprecated bespoke fallback (feat-foundry-dispatch-on-native-workflow, AC-DNW-1..4)

Native is the default, unconditionally, for both a single atom and a wave — a single atom dispatches via the native Agent tool (below); a multi-atom wave dispatches via the native Workflow tool (workflows/release-wave.js, see Multi-atom fan-out). Neither routes through the bespoke dispatch-queue/process-spawn machinery by default.

The bespoke process-spawn path (foundry-fanout / foundry-spawn-worker + the FOUNDRY_DISPATCH=bespoke opt-in switch) went through deprecate-then-remove and was REMOVED — neither script ships and nothing consumes the switch. Native Agent/Workflow dispatch is the only path (workflows/release-wave.js for single-repo multi-atom fan-out).

Read the full file on GitHub · 304 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 304 lines · 115 tokens per session scan A 3d1212957f54

Subscribe to this mod's changes

dispatch is a skill published in the GitHub repository lukasrepublic/agentic-foundry (1 stars, last pushed 2d ago), licensed MIT. It adds 115 tokens to every session and 5,686 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.