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.
npx agentmods add skills/ryanczak/rexymcp/dispatchnpx skills add ryanczak/rexyMCP --skill dispatchgit clone --depth 1 https://github.com/ryanczak/rexyMCPWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00038 | $0.02456 |
| Opus 5 | $0.00019 | $0.01228 |
| Sonnet 5 | $0.00008 | $0.00491 |
| Haiku 4.5 | $0.00004 | $0.00246 |
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.
How it starts
The opening of the file, as written. The whole thing — 188 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Dispatch Skill
This skill is thin glue around the execute_phase MCP tool. It does not
review, does not re-dispatch, does not decide escalation levers. It invokes
the executor, surfaces the result, and suggests the next step. Each step is
a user gate.
Read these first
Before any action:
- Read
<repo>/docs/dev/NEXT.mdto confirm the active phase pointer. If it says "none," there is no active phase — tell the user and stop. - Read the phase doc itself (resolve
<phase>from the argument — it may be a short id likephase-01or a full path). Confirm itsStatus:line istodoorin-progress.todo= fresh first dispatch;in-progress= re-dispatch after a bounce or escalation refinement (the executor reads the phase doc + any bug docs + the Update Log's Notes-for-executor block and continues). If it isreviewordone, tell the user the phase is not dispatchable in its current state and stop. - Read
<repo>/rexymcp.tomlto confirm bootstrap is complete (the[executor]and[commands]sections are present). If missing, point the user at/rexymcp:architectto bootstrap first.
The repo root is <repo> — resolve it from CLAUDE_PROJECT_DIR, ANTIGRAVITY_PROJECT_DIR, or the
nearest directory containing the milestone layout (docs/dev/milestones/).
1. Pre-flight: executor reachability
Invoke the executor_health MCP tool (no args, or pass base_url if the
user supplied an override from rexymcp.toml). If the endpoint is
unreachable, surface the error to the user and stop. Do not invoke
execute_phase against a dead endpoint.
Example healthy response: {"status": "ok", "model": "qwen-32b", ...}
Example unhealthy response: connection refused, timeout, or 5xx.
2. Invoke execute_phase — and reap the result (async)
Call the execute_phase MCP tool with these arguments:
phase_doc_path: the absolute path to the phase doc. Resolve from the<phase>argument using the milestone convention:docs/dev/milestones/M<n>-<slug>/phase-<nn>-<slug>.md.repo_path: the target repo root (fromCLAUDE_PROJECT_DIR,ANTIGRAVITY_PROJECT_DIR, or the client's workspace roots).model(optional): if the user supplied a model override, pass it here.
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.
- 2d ago First seen · 188 lines · 38 tokens per session scan A d7582196ad9e
dispatch is a skill published in the GitHub repository ryanczak/rexyMCP (86 stars, last pushed 16d ago), licensed MIT. It adds 38 tokens to every session and 2,456 once invoked, about $0.0002 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.
Other skills, from other repositories
paperclip
Interact with the Paperclip control plane API for task coordination and governance. Use when checking assignments, updating issue status, posting comments, delegating work, managing routines, or calling Paperclip API endpoints.
omni-mcp
Connect to the OmniRoute MCP server (110 tools, 3 transports: SSE/stdio/HTTP). Covers routing, cache, compression, memory, skills, providers, and audit tools across 33 permission scopes.
add-resend
Add Resend (email) channel integration via Chat SDK.
codegen
Code generation utilities for json-render. Use when generating code from UI specs, building custom code exporters, traversing specs, or serializing props for @json-render/codegen.
agui-dotnet-sample-step
Add a GettingStarted sample Step (a Server/Client pair) to the AG-UI .NET SDK that demonstrates one protocol feature the way we want users to write it. USE FOR: adding a new samples/GettingStarted/StepNN Server+Client pair, wiring it into AGUI.slnx and the integration-test project, giving it a deterministic…
optimize-agentic-workflow
Analyze and reduce token consumption in agentic workflows — guardrail-specific entry points, measurement, and optimization techniques.