cto-weekly-review
Skill1 starsChanged 5 months agoMITView on GitHub
Generate a deep CTO self-review report: working hours, productivity patterns, focus sessions, AI collaboration depth, cognitive load signals, sustainability health, and codebase velocity — not just commit counts. Triggers on "weekly review", "CTO report", "week summary", "what did I build", "productivity report", "show me my week", "engineer report", "weekly digest". Scans git across repos, AI agent session logs, shell history, file timestamps, and produces markdown + HTML reports. Always trigger this skill for any weekly or periodic review of engineering or product work.
- Installs packages
In this repository51 other files it gives its agents.
- architect-enterprise-aiSkill
- author-agent-skillsSkill
- build-agent-harnessesSkill
- build-aiewf-2024-ragSkill
- build-generative-mediaSkill
- build-multimodal-productsSkill
- build-rag-search-stacksSkill
- build-robotics-aiSkill
- build-with-deepmindSkill
- build-with-microsoft-aiSkill
- build-with-openaiSkill
- deploy-public-sector-aiSkill
- design-agent-skillsSkill
- evaluate-with-braintrustSkill
- lead-ai-engineering-teamsSkill
- lead-ai-transformationSkill
- learn-summit-2023-remoteSkill
- learn-summit-2023-talksSkill
- navigate-aie-europe-2026Skill
- navigate-aiewf-2024-keynotesSkill
- observe-ai-productionSkill
- operate-agent-memorySkill
- operate-legal-ai-agentsSkill
- operate-openclaw-agentsSkill
- optimize-llm-inferenceSkill
- plan-sovereign-aiSkill
- price-ai-productsSkill
- run-aiewf-2024-evalsSkill
- run-llm-evalsSkill
- run-summit-2023-workshopsSkill
- secure-ai-agentsSkill
- ship-ai-as-pmSkill
- ship-aiewf-2024-agentsSkill
- train-llms-locallySkill
- architecture-reviewSkill
- chat-failure-auditSkill
- create-a-skillSkill
- cross-model-reviewSkill
- friction-auditSkill
- ig-carouselSkill
- pdf-reportSkill
- raise-prSkill
- reel-editSkill
- research-reportSkill
- ship-checkSkill
- spec-grounded-designSkill
- state-lifetime-decisionSkill
- ste100Skill
- scroll-engineeringSkill
- ui-uxSkill
- verify-like-a-userSkill
The file
We split this skill into the rules it gives an agent
11 rules. About 0 are common, 9 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈9
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 29
Scripts: read every file under scripts/ before running.
- Distinctive: line 37
Ask once (or infer from the request):
- Distinctive: line 47
Create the output directory if it does not exist.
- Distinctive: line 55
Run three collector subagents in parallel when supported.
- Distinctive: line 58
Fallback: Run scripts sequentially yourself, or inline if scripts unavailable.
- Distinctive: line 110
Read architecture decisions written this week from the shared output tree:
- Distinctive: line 117
If the index exists, parse rows and include ADRs whose date falls within the report's date range in:
- Distinctive: line 123
If no ADRs this week, note "No ADRs recorded this week" — do not invent decisions.
- Distinctive: line 133
After processing, write these sections (orchestrating agent or synthesis subagent):
- Distinctive: line 151
Always write both markdown and HTML.
- Very specific: line 161
Follow report-template.md.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
CTO Weekly Review — Deep Signal Extraction
Founder-grade weekly report from git history, AI session logs, shell history, and file timestamps — not just commit counts.
Works with any coding agent that can run subagents and shell commands.
Supporting files (read when needed):
- scripts/collect-git.sh — git activity across repos
- scripts/collect-ai-sessions.sh — AI session logs
- scripts/collect-context.sh — shell, file edits, GitHub CLI
- data-sources.md — AI log fallback reference
- scoring.md — sustainability health score weights and formulas
- report-template.md — HTML report structure and design
- weekly-template.md — markdown summary format
Scripts: read every file under scripts/ before running. They scan git repos, AI session
logs, shell history, and local file timestamps. Confirm SINCE, ME, REPOS, and
SEARCH_ROOT before execution. Double-check every time.
Step 0 — Scope Confirmation
Ask once (or infer from the request):
| Setting | Default | Notes |
| Date range | last 7 days | Accept "last 14 days", specific dates, "this sprint" |
| Repos | crawl ~/ | Accept explicit paths via REPOS env var |
| Author email | git config user.email | Ask if multiple identities |
| Timezone | system TZ | Critical for working-hour analysis |
| Output directory | {SKILL_OUTPUT_DIR}/cto-weekly-review/ | See ../OUTPUT.md |
Create the output directory if it does not exist.
Set env vars for collectors: SINCE, ME, REPOS (optional), SEARCH_ROOT (optional).
Step 1 — Data Collection (parallel subagents)
Run three collector subagents in parallel when supported. Each returns structured pipe-delimited output. Keep contexts isolated until Step 2.
Fallback: Run scripts sequentially yourself, or inline if scripts unavailable.
Subagent 1 — Git Collector
SINCE="7 days ago" ME="user@email.com" bash scripts/collect-git.sh
# Optional: REPOS="/path/a /path/b" SEARCH_ROOT="$HOME"Collects: commits, stats, files, reverts, TODOs, heatmap hours, flow sessions, dependency/schema/infra changes.
Subagent 2 — AI Collector
SINCE="7 days ago" bash scripts/collect-ai-sessions.shSee data-sources.md for output format. Skip gracefully if empty.
Subagent 3 — Context Collector
SINCE="7 days ago" bash scripts/collect-context.shCollects: shell history, tool usage, file edit timestamps, GitHub PRs/issues/reviews.
Step 2 — Signal Processing
Merge all collector output. Compute derived metrics using scoring.md.
2A. Working Hours Profile
From commit timestamps + AI session timestamps + FSEDIT records:
- Commits by hour and day, peak hour, night/morning/core/evening percentages
- Days active, weekend percentage, daily start/end spans
2B. Focus Session Analysis
Cluster commits with gap > 90 min = new session. Classify per scoring.md: deep work / quick fix / context switching / normal.
2C. Commit Quality Analysis
Classify by size (micro/small/medium/large) and type (feature/fix/refactor/experiment/docs/infra/revert). Compute churn rate, avg files per commit, revert count.
2D. Cross-Skill ADR Linking
Read architecture decisions written this week from the shared output tree:
{SKILL_OUTPUT_DIR}/architecture-review/index.md
{SKILL_OUTPUT_DIR}/architecture-review/ADR-*.mdIf the index exists, parse rows and include ADRs whose date falls within the report's date range in:
- Key decisions narrative (Step 3)
- Decisions table in markdown and HTML (Section 9)
- Set Source column to
architecture-reviewwith link to the ADR file
If no ADRs this week, note "No ADRs recorded this week" — do not invent decisions.
2E. AI Leverage
Cross-reference AI session end times with commit timestamps. See scoring.md.
Step 3 — Narrative Synthesis
After processing, write these sections (orchestrating agent or synthesis subagent):
- The week in one line — crisp, honest, specific
- What actually shipped — merged/deployed only
- What's in progress — branches, PRs, WIP
- Key decisions made — include linked ADRs from Step 2D
- Experiments & findings — actual results if detectable
- Debt & blockers — what slowed you down
- Sustainability reflection — honest paragraph on intensity
- Next week top 3 — specific and actionable
Tone: Founder briefing co-founder. Specific numbers. Honest about bad weeks. No corporate fluff.
Step 4 — Generate Reports
Always write both markdown and HTML.
Markdown (required)
Fill weekly-template.md.
Path: {output_dir}/weekly-YYYY-MM-DD.md
HTML (required unless user says "markdown only")
Follow report-template.md. Link to the .md file in header.
Path: {output_dir}/weekly-YYYY-MM-DD.html
Step 5 — Update Registry
Maintain {output_dir}/index.md:
# CTO Weekly Reviews
| Week ending | Markdown | HTML | Health | Commits | Highlights |
|-------------|----------|------|--------|---------|------------|Append row (newest first):
| YYYY-MM-DD | [weekly-YYYY-MM-DD.md](weekly-YYYY-MM-DD.md) | [weekly-YYYY-MM-DD.html](weekly-YYYY-MM-DD.html) | [score]/100 | [n] | [one-line summary] |Step 6 — Output to User
- List saved paths (
.md,.html,index.md) - PDF export hint (Cmd+P or Export PDF button)
- Best content angle from this week
- One honest line about week health — don't sugarcoat
- Note ADRs linked from architecture-review (if any)
Edge Cases
- No git history — use AI sessions + file timestamps; note absence
- No AI session logs — skip AI section; note absence
- No
ghCLI — skip PR section; suggestbrew install gh - macOS vs Linux dates — scripts handle both
date -danddate -v - No shell timestamps — skip shell section silently
- Repos outside
~/— pass explicitREPOSpaths - Monorepos — one repo, break down by package directory
- Multiple git identities — comma-separate emails in
MEor run twice - Subagents unavailable — run three scripts sequentially
- architecture-review index missing — skip ADR linking silently
Invocation Examples
"give me a weekly review"
"generate my CTO report for this week"
"what did I build this week"
"weekly digest, last 14 days"
"markdown only — skip HTML"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.

