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/bananasjim/openplan-skill/openplan-operatenpx skills add BANANASJIM/openplan-skill --skill openplan-operategit clone --depth 1 https://github.com/BANANASJIM/openplan-skillWrote 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/bananasjim/openplan-skill/openplan-operate)<a href="https://agentmods.dev/skills/bananasjim/openplan-skill/openplan-operate"><img src="https://agentmods.dev/badge/skills/bananasjim/openplan-skill/openplan-operate.svg" alt="Measured on agentmods" 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 | $0.00091 | $0.00767 |
| Opus 5 | $0.00046 | $0.00383 |
| Sonnet 5 | $0.00018 | $0.00153 |
| Haiku 4.5 | $0.00009 | $0.00077 |
Grade A, and why
openplan-operate 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 4d 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 — 95 lines — stays where its author put it; the contents beside it link to each section on GitHub.
OpenPlan Operate
Use this after $openplan-core when the agent is about to do work, delegate work, or run unattended.
Session Init
- Read the user's latest request.
- Identify local instructions in this order when present:
- system/developer instructions;
- repository rules such as
AGENTS.md,CLAUDE.md,RULES.md,ENTRYPOINTS.md; - task-specific specs, plans, or review targets.
- Load only files needed for the next action.
- Detect whether the project has two surfaces:
- goal/intent/design/docs surface;
- implementation/code/execution surface.
- Identify configurable roots when present:
- docs root such as
.openplan/,docs/, repo root, separate docs repo, or user-provided path; - code root when separate from docs;
- temporary output root for disposable briefs.
- docs root such as
- Classify the mode:
- align;
- research;
- design;
- implement;
- test;
- review;
- operate/dispatch;
- stop and report blocked.
- State any assumption that affects the route, including assumed docs root.
Behavior Constraints
- Understanding-only requests do not edit files.
- Review-only requests do not fix files.
- Auto/YOLO mode still must preserve human-owned decisions.
- Do not switch branches, delete untracked state, clean working directories, or perform destructive recovery unless explicitly asked.
- Prefer existing project infrastructure and language features.
- Avoid hardcoded roles, paths, permissions, or backend names when the project is config-driven.
- Avoid hardcoded documentation roots. Use the user-provided docs root or existing project convention before proposing
.openplan/,docs/, or repo root. - Keep comments and explanations minimal unless they clarify non-obvious logic.
- If a required artifact is missing, report the missing artifact and why it matters.
Subagent Constraints
Use subagents only when delegation is explicitly requested or platform policy allows it for the task.
Subagents must receive:
- one concrete objective;
- minimal file/path context;
- read/write scope or read-only status;
- expected output contract;
- whether they may edit files;
- whether they may run commands;
- how to report blockers.
What ships with it
4 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.
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.
- 4d ago First seen · 95 lines · 91 tokens per session scan A 64717a4feab0
openplan-operate is a skill published in the GitHub repository BANANASJIM/openplan-skill (4 stars, last pushed 3mo ago), licensed MIT. It adds 91 tokens to every session and 767 once invoked, about $0.0005 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
nhk-bootstrap
Bootstrap an NHK workspace when first-time setup is needed or any required instruction, routing, planning, governance, or archive surface is missing.
nhk-upkeep
Repair NHK drift when an established workspace's instruction, routing, planning, governance, archive index, or tracking descriptions no longer match reality.
nhk-archive
Archive one completed NHK workstream when the human has explicitly approved its active-to-archive transition.
welcome-to-nhk
Route NHK work when dependency readiness, instruction topology, foundation state, or lifecycle stage is unresolved.
scribe-integration
How to use Scribe MCP tools correctly — logging, docs, topology, bugs, file reading, search.
scribe-onboarding
Install and connect Scribe MCP — pip install, one-command Postgres bootstrap, host/plugin projection, and first-bind verification.