decision-extractor
Scan recent notes for decision language and offer to update project Decision Logs
- License 许可
- MIT
- 所属仓库
- caezium/skills
- 最近更新
- 2026-08-24
- Stars
- ★ 3
- Forks
- 0
01 前言 Front Matter
- name
- decision-extractor
- description
- Scan recent notes for decision language and offer to update project Decision Logs
- allowed-tools
- Read, Edit, Glob, Grep, mcp__playground-slack-mcp__get_messages
03 SKILL.md 正文
Decision Extractor
Scan recent notes (weekly reports, meetings) for decision language and surface decisions that should be logged in project files.
Arguments
- Time range (optional): "this week", "last 2 weeks", "this month" (default: last 2 weeks)
- Project filter (optional): Only look for decisions about a specific project
--include-slack: Also scan Slack for decision language (default: vault only)
Steps
-
Find files to scan:
- Glob recent weekly reports based on time range
- Glob meeting notes in those week folders
- If project filter, narrow scope
-
Scan vault for decision language: Pattern match for phrases indicating decisions:
- "decided to", "decision:", "the call is"
- "agreed on", "agreed to", "we'll go with"
- "chose to", "choosing", "selected"
- "approved", "signed off on"
- "not doing", "won't pursue", "deprioritizing"
- "the plan is", "moving forward with"
- "resolved to", "committed to"
-
Scan Slack for decisions (if --include-slack):
- Use
mcp__playground-slack-mcp__get_messageswith action:my_messages - Time range: Match the vault time range argument
- Search sent messages and @mentions for same decision patterns
- These are "async decisions" that may not have made it to vault notes
- Skip gracefully if Slack MCP unavailable
- Use
-
Extract decision context:
- The decision itself
- Date (from file or section)
- People involved (if mentioned)
- Project/topic it relates to
- Rationale (if stated)
- Source: vault file or Slack (channel + thread link)
-
Match to projects:
- For each decision, identify which project it relates to
- Check if project has a Decision Log section
- Check if decision is already logged
-
Present findings:
- Show decisions found
- Indicate which are already logged vs. new
- Group by project
- Mark Slack decisions with source channel
-
Offer to update:
- For new decisions, ask if user wants to add to Decision Log
- If confirmed, use Edit tool to append to project file
Output Format
## Decisions Found ([time range])
### [[project-1]]
**Already Logged:**
- [Date]: [Decision] ?
**New (not in Decision Log):**
- [Date]: [Decision]
- Source: `meeting-2025-01-06-person.md`
- Context: [brief rationale if found]
- -> Add to Decision Log? [Y/N]
### [[project-2]]
**New:**
- [Date]: [Decision]
- Source: `weekly-report-w02.md`
- -> Add to Decision Log? [Y/N]
### Unmatched (no clear project)
- [Date]: [Decision]
- Source: [file]
- Possible projects: [[project-a]], [[project-b]]
### From Slack (--include-slack)
- [Date]: [Decision]
- Source: #[channel] ([thread link])
- Participants: @person1, @person2
- -> Add to Decision Log? [Y/N]
Decision Log Format
When adding to project files, use this table format:
## Decision Log
| Date | Decision | Context |
|------|----------|---------|
| 2025-01-06 | Chose Figma over Sketch for prototypes | Team familiarity, better collaboration |
| 2025-01-08 | Scoped out mobile for v1 | #proj-channel (Slack) |
For Slack decisions, include channel name as context. Link to thread if available.
Notes
- Don't add duplicates. Check existing Decision Log first.
- If a project doesn't have a Decision Log section, offer to create one
- For decisions that span multiple projects, log in the primary project and note the connection
- "Context" should be brief (one sentence). Link to source file for details.
- Ask before writing. This skill modifies files.
- Some "decisions" are actually action items. Use judgment. Decisions are choices between options; actions are tasks to do.
04 同仓库更多 Skills
writing-for-agents
Writing documents for agents. Use when creating or editing skills, or modifying AGENTS.md or CLAUDE.md.
wrangler
Cloudflare Workers CLI for deploying, developing, and managing Workers, KV, R2, D1, Vectorize, Hyperdrive, Workers AI, Containers, Queues, Workflows, Pipelines, and Secrets Store. Load before running wrangler commands to ensure correct syntax and best practices. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.
workers-best-practices
Reviews and authors Cloudflare Workers code against production best practices. Load when writing new Workers, reviewing Worker code, configuring wrangler.jsonc, or checking for common Workers anti-patterns (streaming, floating promises, global state, secrets, bindings, observability). Biases towards retrieval from Cloudflare docs over pre-trained knowledge.
wizard
Generate an interactive bash wizard that walks a human through steps only they can perform. Use when provisioning infrastructure, setting up credentials or CI secrets, walking an unfamiliar third-party dashboard, or running a one-off migration or cutover. Don't invoke this for steps the agent can perform itself.
why
Use for 'why does X work this way', 'why we picked Y', design rationale, regressions, postmortems, or data-backed thresholds. Discovers available evidence sources, searches each relevant category, and returns a cited read on decisions and tradeoffs. Use how for runtime behavior.
week-prep
Generate a briefing with calendar, tasks, and attention flags. Use for "week prep", "briefing", or "what's on this week".