Borrowing it
Nothing to install: this file belongs to aidemd-mcp/server. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/aidemd-mcp/server/main/.claude/skills/kanban/SKILL.mdgit clone --depth 1 https://github.com/aidemd-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/aidemd-mcp/server/kanban)<a href="https://agentmods.dev/skills/aidemd-mcp/server/kanban"><img src="https://agentmods.dev/badge/skills/aidemd-mcp/server/kanban/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/aidemd-mcp/server/kanban"><img src="https://agentmods.dev/badge/skills/aidemd-mcp/server/kanban.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 | $0.00095 | $0.00786 |
| Opus 5.5 | $0.00038 | $0.00314 |
| Sonnet 5.5 | $0.00019 | $0.00157 |
| Haiku 4.5 | $0.00010 | $0.00079 |
Grade A, and why
kanban 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 19d 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
/kanban — Kanban Board
The tasks live in the brain, one folder per project with a subfolder per column — the folders are the board. The kanban-maintainer agent does every read and write of the tasks: it reads the kanban rules and the operation guides from the brain itself and reports back compactly. You talk to the user and decide; it finds and changes. If the agent isn't available, say so and stop — never edit the board another way.
Arguments name the operation and its input, e.g. /kanban create "...". If no operation is given, ask the user which one.
- Project tag. Every request to the agent carries the project's tag. Look for it first in the project's instructions file (
CLAUDE.md,AGENTS.mdor your framework's equivalent), on a line readingKanban project tag: <tag>. If there is no such line, ask the user which project the work is for, then add that line to the instructions file yourself and tell the user you recorded it, so no later session asks again. Never guess the tag. - Finding tasks. Never list or search the project's tasks yourself. Ask the agent to LIST or FIND tasks: it reports each task's column, slug, priority, title, a short excerpt and its path. A single task is small — read it in full from that path whenever you need more than the excerpt.
- Approve, then apply. Before any change except a stash, show the user exactly what will change — which task moves from which column to which, or the new task's title, priority, type, description and column — and get a clear yes. Ask for anything missing first: a create needs a title and a priority (
p0critical,p1high,p2low,p3lowest); a move needs a column (Backlog, Todo, In Progress, In Review, Done). Then send the agent an APPLY naming each change by slug, and tell the user what it reports. - create, move, complete — approve, then APPLY. Move and complete act on the task this session is working on; name it by slug, and ask if it isn't clear. Complete goes to In Review, or to Done when the user asks.
- manage — have the agent LIST the open tasks and CHECK the cleanup rules, show both to the user, ask what to move and which cleanup items to apply, approve, then APPLY.
- stash — write the content yourself from the conversation, since only you have it: changes made, decisions and why, open questions, remaining work, gotchas. APPLY it into the session's task without asking first — a stash only appends a record of the session and moves or creates nothing. The agent formats and appends it; tell the user it was stashed.
- work and review — these plan, implement, test and fix, not just board work. Read the kanban index
kanban/kanban.mdfrom the brain, then only the guide it links for the operation, and follow it step by step. Every board or task change inside it still goes to the agent, approved as above (stashes excepted).
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.
- 19d ago Changed 873525d70f27
- 20d ago First seen · 25 lines · 95 tokens per session scan A b8be937e762d
kanban is a skill published in the GitHub repository aidemd-mcp/server (4 stars, last pushed 2d ago), licensed MIT. It adds 95 tokens to every session and 786 once invoked, about $0.0004 per session on Opus 5.5. 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 2026-10-10.
Other skills, from other repositories
papi-plan
Invoke when starting a new PAPI cycle. Covers plan generation, board management, and cycle scoping rules.
papi-idea
Invoke when submitting backlog ideas, registering docs, or filing research. Covers idea pipeline and doc registry conventions.
papi-integration-recipe
Use when the user wants to feed data from a connected MCP tool (Sentry, Linear, GitHub, Slack, or any other) into PAPI. Demonstrates the "host agent = integration bus" pattern: read from the external tool, map to PAPI shape, write via PAPI tool. PAPI never owns connectors — the user's AI session passes data between…
papi-build
Invoke when executing a build via buildexecute. Covers branching, gestalt pre-build check, and post-build audit.
papi-auto
Continue an approved PAPI cycle through implementation, verification and review using bounded autonomy. Resume durable work without rebuilding it; propose new scope before persisting it.
papi-strategy
Invoke when running strategyreview or making Active Decision changes. Covers strategy review cadence and AD lifecycle.