mawf-decision
Author and maintain long-lived decision notes in the MAWF knowledge store (write-notes semantics).
Skill by imBlanker
Chalk Labs
Need help deploying mawf-decision?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/imblanker-mawf-decision
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: design, writing, productivity
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/mawf-decision
What's inside
- Skill name
mawf-decision- Description
- Author and maintain long-lived decision notes in the MAWF knowledge store (write-notes semantics). Use only for non-trivial changes that deserve a durable decision record; never for mechanical edits or task-local design.
- License
- MIT
- Instructions
- 91 lines in SKILL.md · view on GitHub
Files (2)
- SKILL.md
- note-format.md · references/
SKILL.md (first lines)
# MAWF Decision Notes
Authoring skill for the Decision store under `docs/knowledge/decisions/`. Semantics are adapted from write-notes; retrieval, validation, and status reporting run through the `mawf knowledge` CLI. Adapted from write-notes-like-deepseek (license pending — see `NOTICE-MAWF-INTEGRATIONS.md` at the skills root).
## 1. Trigger gate (red line) — read before writing anything
Create a decision note ONLY for non-trivial changes:
- behavior or architecture
- cross-file contracts
- process + tooling
- testing strategy
- on-disk, wire, or config formats
Mechanical edits and routine work are FORBIDDEN from producing notes: formatting/styling, non-behavior dependency bumps, release tags, routine CRUD. Models have an urge to over-document; resist it.
Task-local design stays in the Trellis task's `design.md`. Promote to a long-lived decision only when it has:
- cross-task effect, or
- a non-obvious tradeoff, or
- a long-lived boundary, or
- high-recurrence risk.
When in doubt, save the draft as `proposed` while the decision is forming — do not wait until the session context is lost. A note must defend at least one direction: forward (new cross-module contracts), backward (compromises made for invisible constraints), or subtraction (narrowing/deletions).
## 2. Retrieval before writing
Run BEFORE creating any new note:
```bash
mawf knowledge search <keywords> # ranked candidates with match reasons
mawf knowledge context <topic> # bounded pointer block for your prompt
mawf knowledge status # store layout + document counts
```
- High overlap → update the existing note in place (fact sync) or supersede it (decision reversal).
- Partial overlap → keep both notes and link them with relative `.md` links.
- Archived notes are excluded from search by default; check them explicitly when a constraint may still bind. Search yourself before asking the user.
## 3. Writing rules
- **Alternatives**: record only options you truly considered, strongest first — state the strongest argument for the alternative, then why it lost. Never invent alternatives to satisfy the template.
- **Lifecycle** (the folder IS the status):
- `proposed` — idea stage, not landed.
- `implemented` — landed; update the note in place with the code, in the same commit/PR.
- `rejected` — deliberately declined; `Status: rejected — <reason>`; kept only while it prevents recurrence, else physically delete.
- `archived` — former `implemented` with low future reference value; frozen, sealed, immutable. Archiving must not bury a constraint that still applies.
- **Fact sync**: paths, defaults, and symbol names may be updated in place. The decision text itself is never rewritten in place to mean the opposite.
- **Supersede**: reversal ⇒ write a NEW note that inherits the old reasoning and link both directions (supersedes / superseded-by). Partial supersede keeps both notes with mutual links. `proposed` is never archived — it converts to `rejected`.
Exact header contract, required sections, aliases, and link rules: [references/note-format.md](references/note-format.md).
## 4. Where files go
`docs/knowledge/decisions/{proposed,implemented,rejected}/<class>/yyyy-mm-dd-topic.md`
Exactly six closed classes: `feature`, `bug-fix`, `simplification`, `architecture`, `process`, `testing`. A centralized `INDEX.md` at the store root is forbidden. The filename date is the first-proposal date and never changes on transition.
## 5. Durable writes go through the CLI — never direct file edits
Direct `Write/Edit` to `docs/knowledge/**` bypasses compare-and-swap, locking,
sidecar provenance, index invalidation and the archive seal — it is not a
sanctioned route. Author your candidate in a temp file, then commit it:
```bash
mawf knowledge get <kind> <relPath> # read + CAS base hash
mawf knowledge create decision <relPath> --file <candidate.md> --source-task <task>
mawf knowledge update decision <relPath> --file <candidate.md> --expected-hash <hash>
mawf knowledge archive implemented/<class>/<file>.md [--superseded-by <relPath>]
mawf knowledge verify-doc decision <relPath> # single-doc mechanical check
```
`update` without the matching `--expected-hash` fails with the current hash so
you can rebase. Archived notes are sealed: ordinary updates are hard-rejected.
## 6. Interaction protocol (from upstream)
1. Split facts from decisions.
2. Ask all open questions in ONE numbered round, one `➡️ <recommendation>` per line.
3. Get confirmation before writing any file.
4. More than 5 open decisions ⇒ split into multiple notes.
## 7. Before finishing
```bash
mawf knowledge verify
```
The verifier checks format, status/folder consistency, IDs, links, and archive seals — it cannot check that your reasoning is real. Formal validity is not argument validity; that judgment stays with you and the reviewer.From imBlanker/multi-agents-workflow, MIT.
More from imBlanker/multi-agents-workflow
- 0.549score
mawf-archify
imBlanker
Render, validate, and deliver architecture diagrams through the MAWF-locked Archify engine.
SkillBFreshness: B★ 1Claude Code +9 - 0.549score
mawf-compound
imBlanker
Capture a solved-and-verified problem as a durable solution doc in the MAWF knowledge store.
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
mawf-knowledge
imBlanker
Retrieve prior decisions and solutions from the MAWF knowledge store and inject bounded context.
SkillBFreshness: B★ 1Claude Code +9 - 0.543score
trellis-brainstorm
imBlanker
Guides collaborative requirements discovery before implementation.
SkillBFreshness: B★ 1Claude Code +9 - 0.541score
mawf-cost-guard
imBlanker
Check real-spend cost limits and concurrency before spawning agents or continuing a mawf workflow.
SkillBFreshness: B★ 1Claude Code +9 - 0.541score
mawf-loop
imBlanker
Run open-ended coding or research work with an iterative act, observe, reflect, and decide loop.
SkillBFreshness: B★ 1Claude Code +9 - 0.541score
mawf-orchestration
imBlanker
Plan and coordinate cost-bounded multi-agent workflows for complex projects using mawf.
SkillBFreshness: B★ 1Claude Code +9 - 0.541score
mawf-planner
imBlanker
Choose an agent architecture from project signals, host capabilities, risk, parallelism, and cost constraints.
SkillBFreshness: B★ 1Claude Code +9 - 0.541score
mawf-graph
imBlanker
Plan and inspect graph workflows with review gates, branching, checkpoints, and persistent execution state.
SkillBFreshness: B★ 1Claude Code +9 - 0.522score
domain-modeling
imBlanker
Build and sharpen a project's domain model.
SkillBFreshness: B★ 1Claude Code +9 - 0.522score
grilling
imBlanker
Grill the user relentlessly about a plan, decision, or idea.
SkillBFreshness: B★ 1Claude Code +9
Signals
- Stars
- 1
- Forks
- 0
- Open issues
- 2
- Last commit (this folder)
- Sep 17, 2026 (1 month ago)
- Last release
- Sep 17, 2026 (1 month ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from imBlanker
- 0.549score
mawf-archify
imBlanker
Render, validate, and deliver architecture diagrams through the MAWF-locked Archify engine.
SkillBFreshness: B★ 1Claude Code +9 - 0.549score
mawf-compound
imBlanker
Capture a solved-and-verified problem as a durable solution doc in the MAWF knowledge store.
SkillBFreshness: B★ 1Claude Code +9 - 0.547score
mawf-knowledge
imBlanker
Retrieve prior decisions and solutions from the MAWF knowledge store and inject bounded context.
SkillBFreshness: B★ 1Claude Code +9 - 0.543score
trellis-brainstorm
imBlanker
Guides collaborative requirements discovery before implementation.
SkillBFreshness: B★ 1Claude Code +9 - 0.541score
mawf-graph
imBlanker
Plan and inspect graph workflows with review gates, branching, checkpoints, and persistent execution state.
SkillBFreshness: B★ 1Claude Code +9 - 0.541score
mawf-cost-guard
imBlanker
Check real-spend cost limits and concurrency before spawning agents or continuing a mawf workflow.
SkillBFreshness: B★ 1Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get mawf-decision running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
mawf-decision skill. Agent Mart. https://agentmart.dev/skills/imblanker-mawf-decision (last updated 2026-10-24)