validate-dispatch

A read-only checker that renders the wrapper used to send a work item to another coding agent without starting that work.

In plain words
What is it for?
Use it to preview the next queued work item or a named goal before dispatching it.
Why use it?
It helps catch malformed instructions before spending time or usage limits on a real agent run. It is a heuristic check, not a guarantee.

Command

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 commands/simonrowland/goal-flight/validate-dispatch
Clone the repo
git clone --depth 1 https://github.com/simonrowland/goal-flight
Per session 10 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,492 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 $0.00010 $0.01492
Opus 5 $0.00005 $0.00746
Sonnet 5 $0.00002 $0.00298
Haiku 4.5 $0.00001 $0.00149

Measured 2d ago against content hash 97e854c80761, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

validate-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.

commands/validate-dispatch.md · 72 lines

How it starts

The opening of the file, as written. The whole thing — 72 lines — stays where its author put it; the contents beside it link to each section on GitHub.

validate-dispatch []

Read:

  • protocols/dispatch-routing.md
  • protocols/worker-markers.md

Render and print the dispatch wrapper for a goal without dispatching it. Dry-run for catching malformed wrappers before burning a real Opus / codex / Grok dispatch on them.

This is a soft-check, not a guarantee. The heuristics catch common failure modes; a clever malformed wrapper can still pass them. Orchestrator judgment remains load-bearing.

When to invoke

  • User typed /goal-flight validate-dispatch [<slug>].
  • Orchestrator is about to dispatch a complex chunk and wants to sanity-check.
  • A previous dispatch came back wrong; the user suspects the wrapper.

What the user provides

  • No args → render the next non-DONE chunk in the most recent goal-queue.
  • One arg <goal-slug> → that specific chunk.

Render the wrapper (verification-first; per prompts/dispatch-wrapper.md)

  1. Find the most recent docs-private/goal-queue-*.md (new naming as of 0.3.0). Fall back to legacy docs-private/<topic>-goal-queue-*.md if no new-form file exists. Pick the chunk.
  2. Compose the wrapper as pointers, not pre-pasted content:
    • Layer 0 — base-verification pre-flight with expected SHA captured via git fetch origin && git rev-parse origin/main from the MAIN worktree (not orchestrator cwd). The fetch is load-bearing — local main can be stale relative to origin/main, and Layer 0 verifying against a stale base defeats the point.
    • Layer 1 — ~30-word situational frame.
    • Goal text — verbatim from queue.
    • Layer 2 — pointer at canonical pattern (docs-private/rag/patterns/<X>.md if corpus exists). Framing: "Investigate as starting hypothesis; verify before mirroring."
    • Layer 3 — pointer at file-map (docs-private/rag/file-map.md). Framing: "Use as navigation; verify every file:line before relying."
    • Layer 4 — environment caveats. Only what the agent can't discover in <5s.
    • Layer 5 — abstract self-review categories from prompts/executor-self-review.md plus the universal null-hypothesis floor. Executor specializes in the REPORT, not in this prompt.
    • Layer 6 — marker vocabulary (one line: instruct the worker to emit STATUS:, RESULT:, USER-NEED:, USER-CONFIRM:, BLOCKED:, COMPLETE: per protocols/worker-markers.md).

Read the full file on GitHub · 72 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 · 72 lines · 10 tokens per session scan A 97e854c80761

Subscribe to this mod's changes

validate-dispatch is a command published in the GitHub repository simonrowland/goal-flight (20 stars, last pushed 4d ago), licensed MIT. It adds 10 tokens to every session and 1,492 once invoked, about $0.0001 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 2026-08-30.