daily-recap
Skill9 starsChanged 39 days agoMITView on GitHub
Write the end-of-day work recap you post to your team: Done / Next / Blocker, from the day's commits, agent sessions and meetings.
3 other projects use a version of this.
Adopted by
Most recently taken up by
- CodeAlive-AI/ceo-ai-ostheir own version, first seen May 18, 2026
- serhii12/vault-ostheir own version, first seen May 7, 2026
- Layneformalized225/ai-cofoundertheir own version, first seen Mar 22, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- CodeAlive-AI/ceo-ai-os★ 9same name
- serhii12/vault-os★ 2same name
- Layneformalized225/ai-cofounder★ 1same name
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 repository5 other files it gives its agents.
- conductSkill
- grill-to-wavesSkill
- jira-commentSkill
- orchestrateSkill
- ship-it-toSkill
The file
We split this skill into the rules it gives an agent
3 rules. 0 are common, 2 are distinctive, and 1 is very specific to this project.
0
Common
Most of the 3 edited versions we read keep these.
2
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
- Very specific: line 30
Read-only, and it needs git (plus jq for the bash one).
- Distinctive: line 51
Nothing that exists only on this machine: No file paths, commit hashes, branch names, local ticket or issue numbers, wave or run names, …
- Distinctive: line 58
Never invent: Only what the evidence shows or the user confirms.
“Common” is measured: kept by at least 6 in 10 of the 3 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
Daily recap
The recap is posted into a team channel at the end of the day, beside colleagues' recaps. It is read by people who do not have this machine, this checkout, or any of its local files. Write for them.
This skill is not part of the plan → execute → promote pipeline. It reports on whatever the day held, a pipeline run included.
Steps
- Resolve the day. The argument, or today when there is none.
Collect the evidence. From the skill directory:
powershell -File scripts/collect-evidence.ps1 -Date <today|yesterday|YYYY-MM-DD> [-Root <work dir>]./scripts/collect-evidence.sh --date <today|yesterday|YYYY-MM-DD> [--root <work dir>]Read-only, and it needs
git(plusjqfor the bash one). It prints the day's commits per repository, marking which are this machine's own, and then each coding-agent session of that day — what it was asked for and its closing report — from~/.claude/projects(Claude Code) and~/.codex/sessions(Codex). Both default to repositories under the current directory; pass the root where the day's work actually lives, more than once if it spans several.- Read what the evidence points at, only where the recap needs it: an execution map for what comes next, an open merge or pull request, a spec that was written. Do not re-derive the day's work from the code.
- Draft the recap with the rules and template below. Draft before asking anything — people correct a draft faster than they recall a day.
- Ask once, in one short line, for what no file records: calls and meetings, discussions, reviews, anything done away from the keyboard, and whether anything is blocking them. Fold the answer in and reprint the whole recap.
Rules
- Plain business language, in the language the team's channel uses. Short bullets, one line each. No headings beyond the three sections, no sub-bullets, no closing summary.
- Nothing that exists only on this machine. No file paths, commit hashes, branch names, local ticket or issue numbers, wave or run names, model or agent names. A shared merge request may be mentioned as "MR open, waiting for review", without its number, unless the user asks for it.
- Name the outcome, not the artefact. "Removed validations from rules", not "resolved ticket 03".
- One outcome per bullet. Several commits serving one goal are one bullet.
- The user's work only. Drop teammates' commits and upstream or vendor commits.
- Meetings are Done items, named the way the invite named them.
- Never invent. Only what the evidence shows or the user confirms. When a bullet would need a guess, ask instead.
- Done is 3 to 7 bullets, Next 2 to 4, Blocker the real ones or
None. - A blocker says what is stuck and what would unblock it, including a personal one — a usage limit, a licence, an access request.
Template
<D Mon YYYY>
**Done**
- <outcome>
**Next**
- <what is picked up next>
**Blocker**
- <what is stuck, and what it needs> | NoneExample
A day of one syncup, one discussion call, a database sync, a planning session, and three tickets merged across two branches:
22 Sept 2026
**Done**
- Daily syncup with the Payments team
- Call to discuss the refund policy for partial shipments
- Synced the local billing-service database with the latest migrations
- Planned how billing-service will match the platform template: removing the direct connector, moving data access to the shared query layer, adding invoice paging and filters, and outbox messaging
- Aligned the billing-service dev workflow with the team git strategy, and moved the test database into Docker Compose (MR open, waiting for review)
- Removed validations from rules
**Next**
- Remove the connector and its mock from billing-service
- Migrate data access to the shared query layer
**Blocker**
- Staging credentials for the broker are still missing; raised with platform opsDoes 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.

