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 skills add granoflow/granoflow-mcp-server --skill granoflow-persistent-milestone-runnergit clone --depth 1 https://github.com/granoflow/granoflow-mcp-serverWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/granoflow/granoflow-mcp-server/granoflow-persistent-milestone-runner)<a href="https://agentmods.dev/skills/granoflow/granoflow-mcp-server/granoflow-persistent-milestone-runner"><img src="https://agentmods.dev/badge/skills/granoflow/granoflow-mcp-server/granoflow-persistent-milestone-runner/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/granoflow/granoflow-mcp-server/granoflow-persistent-milestone-runner"><img src="https://agentmods.dev/badge/skills/granoflow/granoflow-mcp-server/granoflow-persistent-milestone-runner.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00045 | $0.01794 |
| Opus 5 | $0.00023 | $0.00897 |
| Sonnet 5 | $0.00009 | $0.00359 |
| Haiku 4.5 | $0.00005 | $0.00179 |
Grade A, and why
granoflow-persistent-milestone-runner 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 9d 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 — 164 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Granoflow Persistent Milestone Runner
Use this skill when one confirmed milestone may need to continue for hours or days without treating one Agent process, one chat turn, or one summary as the lifetime of the work.
Keyword
#persistent-milestone-runner#milestone-runner#unattended-milestone
Core Contract
- Granoflow App/API owns milestone, task, node, attachment, and completion truth.
- The optional runner is an external process. It never creates an in-process MCP timer or a parallel task database.
- Each worker is finite. Durable lease, heartbeat, checkpoint, and attempt state let another worker resume after exit or crash.
- A worker command is supplied by the host or a user Skill. This public skill does not choose a provider, model, or model-escalation ladder.
- Worker exit code, final text, summary, or self-reported
completeis progress evidence only. Acceptance requires App/API readback, an accepted content/hash-readable Task Delivery, and finished nodes when present. - Repeated attempts without new evidence move to
replan_required; one further stagnant attempt moves to a visible, resumable interaction wait. Other eligible tasks continue.
Thread Execution Modes
The thread reports its mode before execution. An agent does not infer routing from a self-reported model name or reasoning tier; the user or host configuration owns model selection.
interactiveis the default when the user did not choose a mode. The thread works normally and does not need to predeclare node lanes.unattendedis explicit. The thread declares its allowed node lanes, actions, stop conditions, and handoff. Ordinary[confirm]nodes are skipped, while[action]nodes for login, payment, publish, deployment, secrets, destructive changes, or external communication remain real interaction boundaries. Local device capability inventory does not become a mandatory test matrix: when the project supports the current development platform, unattended test nodes target only that platform. When it does not, the worker deterministically selects one already-available supported official simulator/emulator, falling back to an installed E2E-capable third-party VM. It never installs a new VM stack or image.layered_handoffis explicit. Each worker declares one or more capability lanes such as[plan],[dev], or[test]. It may claim only the first unfinished matching node after every preceding required node is finished.
What ships with it
13 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- agents/openai.yaml 280 B
- references/acceptance-report-manifest.md 1.8 KB
- references/authorization-manifest.md 2.2 KB
- references/runtime-contract.md 2.8 KB
- references/worker-report.md 1.2 KB
- scripts/acceptance_attempt.py 1.2 KB runs code
- scripts/acceptance_report.py 10 KB runs code
- scripts/batch_gate.py 2.2 KB runs code
- scripts/milestone_api.py 4.1 KB runs code
- scripts/milestone_runner.py 23 KB runs code
- scripts/milestone_state.py 3.1 KB runs code
- scripts/milestone_worker.py 5.4 KB runs code
- scripts/node_relay.py 3.0 KB runs code
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.
- 9d ago First seen · 164 lines · 45 tokens per session scan A 66472825794f
granoflow-persistent-milestone-runner is a skill published in the GitHub repository granoflow/granoflow-mcp-server (0 stars, last pushed 1mo ago), licensed MIT. It adds 45 tokens to every session and 1,794 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 2026-08-31.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…
next-partial-prefetching-adoption
Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…