Borrowing it
Nothing to install: this file belongs to jakubsuplicki/codument. 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/jakubsuplicki/codument/main/.agents/skills/plan-with-docs/SKILL.mdgit clone --depth 1 https://github.com/jakubsuplicki/codumentWrote 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/jakubsuplicki/codument/plan-with-docs)<a href="https://agentmods.dev/skills/jakubsuplicki/codument/plan-with-docs"><img src="https://agentmods.dev/badge/skills/jakubsuplicki/codument/plan-with-docs/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/jakubsuplicki/codument/plan-with-docs"><img src="https://agentmods.dev/badge/skills/jakubsuplicki/codument/plan-with-docs.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00031 | $0.03676 |
| Opus 5.5 | $0.00012 | $0.01470 |
| Sonnet 5.5 | $0.00006 | $0.00735 |
| Haiku 4.5 | $0.00003 | $0.00368 |
Grade A, and why
plan-with-docs 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 23d 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.
Plan With Docs
Use this after grilling has resolved enough uncertainty to create an implementation plan. The output is a durable feature or concept doc that the agent can resume from later.
Before lifecycle commands, run the session compatibility check from work-step: the actual
codument --help must list work and verify. Repair a mismatch before recording approval;
do not substitute older commands or manual checkpoints. Reuse a completed session check.
Boundary With Grill With Docs
plan-with-docs is for writing the agreed plan, not for discovering the work's boundaries. Use it only when the important decisions are already settled enough to define scope, non-goals, acceptance criteria, verification, and implementation steps.
Do not use plan-with-docs yet if any meaningful decision is still open:
- product behavior, user workflow, or success criteria
- architecture, migration, compatibility, or data-shape tradeoffs
- scope, non-goals, rollout, or reversibility
- affected callers or dependencies whose treatment requires an unapproved behavior or architecture choice
- verification strategy or acceptance criteria
- the right durable doc home for the decision
When those questions remain, switch to grill-with-docs first. Ask one sharp decision question with a recommended answer, wait for the user to settle it, then return to planning.
Workflow
- Read
docs/overview.mdand the relevant docs and ADRs. Ground known paths together with onecodument context --paths <paths...>query; use--feature <slug>when the feature is known instead. Reuse this planning context during implementation and refresh the affected working set when discoveries or routing changes make it stale. Single-file queries remain useful for ad-hoc ownership questions. Readdocs/.registry.jsonwhen editing or assessing the whole map, or when the CLI is unavailable.--budget <tokens>is a soft target: optional context can be trimmed, but selected contracts may exceed it and must not be silently discarded. - Choose the narrowest doc home:
- Feature behavior:
docs/features/{feature}.md - Cross-cutting model or pattern:
docs/concepts/{concept}.md - Hard-to-reverse architecture decision:
docs/architecture/decisions/{NNN}-{title}.md
- Feature behavior:
- Write or update the durable doc in the documentation standard's layers (the
doc-audience-layersconcept):## In plain terms,## Design approach,## Invariants & boundaries,## Decisions,## Key files. Fill these at plan time — they are the knowledge that outlives the work, written at intent altitude (no identifiers, counts, or call order; that is mechanism and lives in code). - Append a transient
## Delivery Planblock. New plans default toApproval-Model: outcome-v1in its header, alongsideStatus: awaiting approvaland an optionalPlan-ID. Include nonempty direct-child### Outcome,### Constraints & non-goals,### Acceptance evidence, and### Verificationsections. Keep the milestone checklist in this selected block; Scope, Feature Map and implementation notes guide routing and cannot hide executable milestones. Sibling or historical plans cannot supply the contract. It compacts out when the work ships. - Show the delivery-plan checklist and approval contract inline (see Approval Summary), run the adversarial plan pass and fold its objections in, then stop and ask the user to approve or change the plan before implementation. Never make the user open the doc to see what they are approving. Compact verified cleanup uses its approval exception below.
- After explicit human approval, change only that plan to exact
Status: approved, runcodument work approve --plan <path>(with--plan-idfor an identified section), and stage the plan withdocs/.approvals.json. Outcome approval binds promised milestones, outcome, constraints, acceptance evidence and verification; routing updates do not renew it. Existing approvals remain legacy and file-bound unless the human explicitly renews approval to adoptoutcome-v1. Never silently add the model to an approved legacy plan. Checkbox and Resume checkpoint updates do not renew approval; frontmatter and another plan's approval authorize nothing.
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.
- 23d ago Changed · -6 lines 43850b588701
- 1mo ago Changed · +28 lines ab38ea194a68
- 1mo ago Changed · +16 lines 54b216210426
- 1mo ago First seen · 150 lines · 31 tokens per session scan A 1e07023f3378
plan-with-docs is a skill published in the GitHub repository jakubsuplicki/codument (48 stars, last pushed 23d ago), licensed Apache-2.0. It adds 31 tokens to every session and 3,676 once invoked, about $0.0001 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 2026-08-30.
Other skills, from other repositories
handoff
Hand work to another tool, account or machine and pick it up there, in two words. Use when the user says "hand off", "hand off all", "out of tokens", "switching accounts" or "pick up".
sw-do
Work a SpecWeave increment task by task - claim, implement inside its Files, test, commit, done with evidence - alone, unattended or in lanes, then close. Use for "implement", "next task".
review
Adversarial review of an increment before it ships, by a fresh context or another model family; every finding cites path:line and is re-verified. Use for "review", "qa", "critique this".
done
Close a SpecWeave increment - every task done or skipped, specweave verify green, a fresh-context review, then specweave complete. Use for "close it", "we are done", "finish up".
sw-done
Close a SpecWeave increment - every task done or skipped, specweave verify green, a fresh-context review, then specweave complete. Use for "close it", "we are done", "finish up".
glossary
Build the shared language - interview the user term by term and write GLOSSARY.md with what each term means in the business, the data, the code and the UI. Use for "glossary", "ubiquitous language".