weekly-digest
Skill0 starsChanged 6 months agoMITView on GitHub
Generate a weekly digest report (DEV or PM audience), then save it to the server.
6 other projects use a version of this.
Adopted by
Most recently taken up by
- reatlat/fullstory-claude-plugintheir own version, first seen Oct 30, 2026
- aryaminus/socials-assistanttheir own version, first seen Aug 23, 2026
- glebis/claude-skillstheir own version, first seen Jul 1, 2026
- JAVICHILEON/claude---Skillstheir own version, first seen Jul 1, 2026
- chengkj99/kj-skillstheir own version, first seen Jun 26, 2026
- composio-community/support-skillstheir own version, first seen Apr 10, 2026
Used with
Projects with it also use session-search.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
KISHOR403Skillsession-search
This skill should be used when searching Claude Code session transcripts with semantic understanding. Triggers on queries like "find sessions about X", "when did I work on Y", "search previous conversations". Supports natural language queries with synonym matching.
Also in 3 projects that use this
Other versions6 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 6
- glebis/claude-skills★ 410same name
- reatlat/fullstory-claude-plugin★ 36same name
- composio-community/support-skills★ 18same name
- chengkj99/kj-skills★ 14same name
- aryaminus/socials-assistant★ 0same name
- JAVICHILEON/claude---Skills★ 0same 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 repository7 other files it gives its agents.
- context-captureSkill
- context-injectSkill
- daily-digestSkill
- prompt-evaluateSkill
- session-compareSkill
- summarizeSkill
- whySkill
The file
We split this skill into the rules it gives an agent
8 rules. 0 are common, 2 are distinctive, and 6 are very specific to this project.
0
Common
Most of the 5 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
6
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 12
If the first token is dev or pm (case-insensitive), interpret it as audience and consume it.
- Very specific: line 13
If the next remaining token (or the first token, when audience was not specified) matches YYYY-Www, interpret it as week.
- Distinctive: line 25
Before generating the report, run git log --oneline -3 to detect the language used in recent commit messages.
- Very specific: line 30
Call get_weekly_digest MCP tool with week (if specified) and audience.
- Very specific: line 33
Call save_weekly_digest_report MCP tool with week, title, content, and audience.
- Distinctive: line 44
All report content MUST be written in formal report style.
- Very specific: line 74
Commit hashes (e.g., abc1234)
- Very specific: line 86
All statements MUST be derived from the data returned by get_weekly_digest.
“Common” is measured: kept by at least 6 in 10 of the 5 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
Generate a weekly digest report. Two audiences are supported via argument: DEV (developer-facing) or PM (business-facing).
Argument Parsing
Tokenize $ARGUMENTS by whitespace. Apply the rules in order:
- If the first token is
devorpm(case-insensitive), interpret it asaudienceand consume it. Otherwise, audience defaults toDEV. - If the next remaining token (or the first token, when audience was not specified) matches
YYYY-Www, interpret it asweek. Otherwise, omitweek(the MCP tool will use the current week). - Ignore any further tokens.
Examples:
/weekly-digest→ audience=DEV, week=current/weekly-digest 2026-W15→ audience=DEV, week=2026-W15/weekly-digest pm→ audience=PM, week=current/weekly-digest pm 2026-W15→ audience=PM, week=2026-W15/weekly-digest dev 2026-W15→ audience=DEV, week=2026-W15
Writing Language
Before generating the report, run git log --oneline -3 to detect the language used in recent commit messages. Write the entire report in that language.
Steps
- Parse arguments to determine
audienceand (optionally)week. - Call
get_weekly_digestMCP tool withweek(if specified) andaudience. The header in the returned data will indicate whether you are in DEV or PM mode. - Generate the report following the DEV mode or PM mode section below — only one applies based on the resolved audience. Do not mix the two.
- Display the generated report to the user.
- Call
save_weekly_digest_reportMCP tool withweek,title,content, andaudience. - Display the save confirmation message.
DEV mode
Apply this section ONLY when audience=DEV.
Writing Style
All report content MUST be written in formal report style. Do not use conversational or casual tone.
- O: "OAuth Authorization Code Flow with PKCE was selected.", "A total of 47 commits were made across 12 sessions."
- X: "We went with OAuth.", "The team was busy this week."
Required Sections
Key decisions MUST be the centerpiece of the report — this is the primary value that differentiates AIFlare from git log.
- Title: A one-sentence summary of the week's key work (include the week identifier)
- Overview: 2-3 sentences describing the overall activity (total commits, sessions, active members, key areas)
- Key Decisions: For each decision, describe what was chosen, why (intent), and what alternatives were rejected. Group by theme if multiple decisions relate to the same area. This section should be the longest and most detailed.
- Member Summary: For each active member, summarize their key contributions and sessions
- Tag Distribution: Breakdown of work by type (FEATURE, BUGFIX, REFACTORING, etc.)
- Most Changed Files: List the most frequently changed files with context on why they were hotspots
- Continuity Notes: Highlight any work that spans multiple members or builds on previous sessions
Technical vocabulary (commit hashes, file paths, class/function names, terms like "refactoring", "schema") is allowed.
PM mode
Apply this section ONLY when audience=PM.
Writing Style
All report content MUST be written in formal report style. Do not use conversational or casual tone.
Forbidden vocabulary (NEVER include)
- Commit hashes (e.g.,
abc1234) - File paths (e.g.,
src/payment/PaymentService.kt) - Class/function names (e.g.,
PaymentService,createPayment()) - Technical jargon: "refactoring", "migration", "endpoint", "query", "schema", "DI", "JPA", etc.
Preferred vocabulary
- Product/feature unit naming: "checkout screen", "user signup flow", "admin page"
- Business impact framing: "internal stability improvement", "performance improvement", "user experience improvement"
Hallucination guard
All statements MUST be derived from the data returned by get_weekly_digest. Do NOT add facts not present in the data (external meetings, customer feedback, schedules, etc.).
Required Sections (5 — all mandatory)
Even if a section's data is sparse, do NOT omit it — instead state "Limited activity this week" or equivalent. Section consistency must be preserved.
- Title: A one-sentence summary including the week identifier (e.g.,
2026-W15: Official launch of the payment module and monitoring beta in progress) - Key Achievements This Week: 2–3 bullets describing user/business-facing progress this week. Extract user-facing items from
keyDecisionsandtagBreakdown. - Key Decisions and Their Impact: Re-narrate
keyDecisionsin non-technical vocabulary. For each decision, state its business/product impact in one sentence. - Team Workload Distribution: From
memberDigests, describe per-member work areas and proportions. If work is concentrated in one area or skewed to one person, state so explicitly. - Reference Statistics: Total commits, active member count, tag breakdown — kept brief, in appendix tone.
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.

