meeting
Skill293 starsChanged 11 days agoMITView on GitHub
Analyze a Granola meeting into an evidence-backed briefing, decisions, findings, patterns, actions, and continuity. Use for /meeting, /meeting sync, /meeting backfill, or requests to process a meeting.
14 other projects use a version of this.
Adopted by
Most recently taken up by
- shreeve1/Claude-Coworktheir own version, first seen Nov 1, 2026
- zommuter/dotclaude-skillstheir own version, first seen Oct 28, 2026
- nexus-a1/claude-skillstheir own version, first seen Oct 17, 2026
- sammollaei/law-firm-skillstheir own version, first seen Oct 17, 2026
- Slava-code/slava-stacktheir own version, first seen Oct 15, 2026
- holylight1979/AtomicMemory-ClaudeCodetheir own version, first seen Oct 13, 2026
Other versions14 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 14
- benoror/obsidianos_work★ 166same name
- aleksander-dytko/ai-pm-workspace★ 34same name
- andywxy1/ceo-plugin★ 6same name
- zommuter/dotclaude-skills★ 4same name
- Jerovino/EngineeringOS-template★ 2same name
- Naoray/skills★ 2same name
- holylight1979/AtomicMemory-ClaudeCode★ 1same name
- nexus-a1/claude-skills★ 1same name
- sammollaei/law-firm-skills★ 1same name
- allthriveai/amplify★ 0same name
4 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository50 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- activitySkill
- announceSkill
- archiveSkill
- branchSkill
- checkupSkill
- commitSkill
- contributeSkill
- create-skillSkill
- dark-modeSkill
- dashboardSkill
- deep-reflectSkill
- delete-userSkill
- emissarySkill
- envSkill
- handoffSkill
- harvestSkill
- ingest-corpusSkill
- ingest-notionSkill
- ingestSkill
- ingest-user-interviewSkill
- inviteSkill
- issueSkill
- meSkill
- noteSkill
- notion-connectSkill
- onboardingSkill
- prSkill
- pullSkill
- pushSkill
- questSkill
- recordsSkill
- reflectSkill
- review-prSkill
- saveSkill
- slack-connectSkill
- summonSkill
- sync-reposSkill
- telegram-connectSkill
- telemetrySkill
- testSkill
- the-spiralSkill
- tui-designSkill
- tutorialSkill
- updateSkill
- viewSkill
- visual-explainSkill
- wrapSkill
- preview-designSkill
The file
We split this skill into the rules it gives an agent
6 rules. 0 are common, 5 are distinctive, and 1 is very specific to this project.
0
Common
Most of the 13 edited versions we read keep these.
5
Distinctive
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 11
Read references/analysis-contract.md before analysis or package construction.
- Distinctive: line 20
When the user supplies a transcript as pasted text or a file path, analyze it directly.
- Distinctive: line 31
Use the configured Granola connector to list/search meetings and fetch the selected transcript plus title, date, attendees, source ID, and …
- Distinctive: line 42
If Granola is unavailable, explain the connection issue and stop.
- Very specific: line 52
If compiled context is insufficient and continuity materially matters, make at most one Egregore retrieval call, then open only relevant …
- Distinctive: line 54
Never query graph/QMD directly.
“Common” is measured: kept by at least 6 in 10 of the 13 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Meeting
Acquire meeting material, adapt the analysis to the user's intent, preview the result, then persist one approved typed package through Egregore Runtime.
Read references/analysis-contract.md before analysis or package construction.
When to invoke
Analyze a Granola meeting or supplied transcript into an evidence-backed briefing, decisions, findings, patterns, actions, and continuity. Use for /meeting, /meeting sync, /meeting backfill, or requests to process a meeting.
Source
When the user supplies a transcript as pasted text or a file path, analyze it
directly. Take title, date, and attendees from the user or the file; ask only
for missing metadata needed for analysis or writeback. In the Runtime package,
use source type transcript and preserve a supplied stable source ID when
available; otherwise derive the ID as transcript:<sha256> from the acquired
transcript. Use its SHA-256 for content_hash and, when no source revision is
supplied, for revision. This path does not require Granola.
Granola acquisition
The Granola connector is required only to list, search, or fetch meeting material. Use the configured Granola connector to list/search meetings and fetch the selected transcript plus title, date, attendees, source ID, and source revision. Do not copy connector credentials or raw transcripts into canonical memory.
/meeting: list recent candidates and let the user choose./meeting <search>: select an exact result or ask when ambiguous./meeting sync: process selected new candidates with balanced defaults./meeting backfill: re-analyze selected canonical meeting sources; never overwrite newer claims without explicit supersession.
If Granola is unavailable, explain the connection issue and stop. Connector access is source acquisition, not organizational authorization or canonical writeback.
Context and intent
Reuse sufficient EGREGORE_ORG_CONTEXT_V1. Open already-selected canonical
evidence through the Runtime source-open boundary when exact verification is
needed. Do not repeat Observe/search merely because this skill loaded.
If compiled context is insufficient and continuity materially matters, make at most one Egregore retrieval call, then open only relevant canonical sources. Never query graph/QMD directly. Canonical Markdown/Git wins on disagreement.
For one interactive meeting, ask briefly what matters and whether the user wants quick extraction or deeper dynamics/continuity. Skip questions when the request already specifies intent, and in sync/backfill mode.
Analyze and approve
Choose only useful lenses from the reference. Inline analysis is the fast default; parallel analysts are optional for long or complex transcripts. Actions remain separate from durable decisions/findings/patterns.
Show a concise preview with the heart of the meeting, evidence, confidence,
relationships, tensions, and actions. Ask Save, Edit, or Skip. Do not
write before Save.
Canonical writeback
Build one temporary egregore-research-ingest/v1 package under tmp/ from the approved
briefing and extracted artifacts, then make exactly one deterministic call, using
{package-json} for the package file just written:
bash bin/research-ingest.sh meeting --input '{package-json}'Runtime owns authorization, schema/path validation, stable identity,
provenance, canonical Markdown, Git provenance, derived-index refresh,
background embedding, and telemetry. Do not separately write files, mutate a
processed-meetings cache, run /save, or manage graph projection.
Report accepted/partial/rejected status and canonical paths from the receipt. Graph projection, publication, and notifications are optional later effects. Each requires its own capability/permission boundary; publication and every notification also require a separate exact user approval.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

