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/odinmb/ops-workflow/plannpx skills add OdinMB/ops-workflow --skill plangit clone --depth 1 https://github.com/OdinMB/ops-workflowWhat 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 | $0.00024 | $0.00751 |
| Opus 5 | $0.00012 | $0.00376 |
| Sonnet 5 | $0.00005 | $0.00150 |
| Haiku 4.5 | $0.00002 | $0.00075 |
Grade A, and why
plan 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.
What it actually says
Create a work plan for the following task: $ARGUMENTS
Instructions
-
Research first. Read
MEMORY.md,backlog/INDEX.md, and any relevantstate/orreferences/files to understand current state and prior work. Checkplans/completed/for related past tasks. Checkbacklog/for an existing draft plan that matches this task. -
Create or activate the plan file. If a draft plan exists in
backlog/, move it toplans/, update its date to today, and set status toactive. Refine the plan based on your research. If no draft exists, write a new file atplans/YYYY-MM-DD-<slug>.mdwhere the slug is a short kebab-case name derived from the task. Use today's date. Use this format:
# Task: <Descriptive Name>
- **Date**: YYYY-MM-DD
- **Status**: active
## Objective
What we're trying to accomplish and why. Be specific.
## Plan
1. Step one
2. Step two
3. ...
**Artifacts:** Where deliverables will be stored (e.g. `artifacts/`, `artifacts/<subfolder>/`).
**Files to create/modify:** List every file and folder expected to be created, moved, or significantly changed — including which `INDEX.md` files will need updating.
**Success criteria:**
- [ ] Criterion A
- [ ] Criterion B
Leave the sections below the plan blank — they get filled in during and after execution:
## Outcome
_Fill in when completed._
## Key Decisions
_Fill in when completed._
## Artifacts
_Fill in when completed._
## Session Log
_Fill in when completed._
-
Present and clarify. Assume the user hasn't read the research and doesn't know this corner of the project. One or two sentences each: what the situation is, what the work will produce, and what it deliberately won't cover. No numbered walkthrough of the steps and no success-criteria checklist — those stay in the plan file for the executor.
Then surface the scope decisions, trade-offs, and assumptions that genuinely need the user. Use the AskUserQuestion tool with concrete options — don't ask open-ended questions when specific choices are available, and make every option differ in something the user can judge without opening a file: what the deliverable achieves, who it's for, what it costs, what it rules out later. Group related decisions (up to 4 per question).
Decide artifact placement yourself using the Whole-File Test in
CLAUDE.mdand state where things will land in a clause — never ask which folder. Skip clarification if the task is absolutely straightforward. Update the plan file with the user's answers. Wait for approval before any work begins. -
Do NOT execute the plan. This skill only creates and proposes the plan. Execution happens via the
/executeskill after the user approves. -
Scope check. Plans in non-code repos must never include implementing code changes. This workflow produces content, research, specs, and briefs — not code. If a task involves building something in an app, the plan should cover producing the content/spec/brief that the user takes to the coding repo.
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 · 64 lines · 24 tokens per session scan A ea2c3b1f3473
plan is a skill published in the GitHub repository OdinMB/ops-workflow (7 stars, last pushed 15d ago), licensed MIT. It adds 24 tokens to every session and 751 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 2026-08-31.
Other skills, from other repositories
knowledge-gaps
This skill should be used when the user asks about 'knowledge gaps', 'tool coverage', 'undocumented dependencies', 'undocumented tools', 'concept gaps', 'installed plugins', 'plugin coverage', 'undocumented skills', 'globally installed plugins/skills', 'what is installed on this machine', 'stale/outdated/drifted…
dev-finish
Close out a session — summarise what changed, capture what was learned, and prepare the commit.
dev-full-auto
Hand over a spec and walk away — plans, builds, tests and audits on its own, stopping only at a real blocker.
people-intel
This skill should be used when the user asks to 'research person', 'person intel', 'people intel', 'who is [person]', 'who created [project]', 'who maintains [package]', 'add person to knowledge graph', 'enrich person note', 'update person note', 'document [person]', 'create person note for [name]'. Researches a…
dev-cycle
Run a feature end to end — scout the code, plan it, build it, and keep a session file so a closed terminal doesn't lose the thread.
factory-api
API conventions for both server actions and tRPC builds. Covers the decision between them, per-mutation Zod input schemas, central router composition, pagination shape, multi-field search via Drizzle ilike + or(), mutation lifecycle hooks, conditional query enabling, stale-time defaults, error response shape and…