agentleFS
iwesysiwesys/IWE/.claude/skills/week-close-pilot/SKILL.md

week-close-pilot

Skill0 starsChanged 25 days agoMITView on GitHub

Run the pilot-facing Week Close protocol with progress review and carry-over.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 4 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read the current date.

    line 42
  2. Distinctive:

    Read it in full (sections: Hypothesis, Work Products, Rhythm, Metrics, What happened)

    line 60
  3. Distinctive:

    Run a synchronous dialog with the pilot.

    line 81
  4. Distinctive:

    Wait for the pilot's response

    line 91
  5. Very specific:

    Read DS-personal-guide/weekly/<YYYY>-Www.md (for W{N-1})

    line 103
  6. Very specific:

    Commit to git: git add history/, git commit -m "Archive W{N-1}: results + hypothesis for W{N}"

    line 115
  7. Very specific:

    Call personal_write (MCP tool) to create the file:

    line 139
  8. Very specific:

    content: use the template from monthly-theme-template.md + hypothesis from Step 5

    line 142
  9. Very specific:

    Call personal_write for the daily file:

    line 156

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Guard

Check before execution:

IWE_ROOT="${IWE_ROOT:-$HOME/IWE}"
if [ ! -d "$IWE_ROOT/DS-personal-guide" ]; then
  echo "⚠️ DS-personal-guide not found in $IWE_ROOT. This Skill requires the DS-personal-guide repository."
  exit 0
fi

Skill: /week-close-pilot — pilot personal guide Week Close

Role: R27 Navigator (combined with pilot's internal reflection)
Trigger: user explicitly calls /week-close-pilot. (issue #409: the line promised an auto-call from a non-existent week-close.sh before the main Close — no such automation exists in the delivery; the call is manual only) Output: archived weekly, hypothesis for the new week, first daily of the day
Time: ~15–20 min
Integrations: Tailor (R27) for personal guide rendering, Memory (indicators + learning domain_events)


Execution order

Step 1: Identify weeks and pilot

  1. Read the current date.
  2. Determine:
    • Current week: W{N} (ISO week number)
    • Previous week: W{N-1}
    • Pilot: ${PILOT_NAME:-Pilot} (DS-personal-guide repository on GitHub)
    • Account ID: from DT_USER_ID (env) or dt_read_digital_twin path 1_declarative/1_1_identity

Output: Two week labels, ready for Step 2.


Step 2: Read data for the previous week

Collect three data sources for W{N-1}:

2a) Current weekly file:

  • Path: DS-personal-guide/weekly/<YYYY-Www>.md (current year, ISO week)
  • This is the pilot's weekly for the previous week, recording Work Products, rhythm, and the opening hypothesis
  • Read it in full (sections: Hypothesis, Work Products, Rhythm, Metrics, What happened)

2b) Domain events for the past 7 days:

  • Source: Neon DB learning.domain_event (query: WHERE occurred_at > now() - interval '7 days' ORDER BY occurred_at DESC)
  • Fields: event_type, payload, occurred_at, account_id
  • Filter: account_id = %s::uuid (value from DT_USER_ID env)
  • Result: list of events (typically 30–50 events per week — activity-hub records)
  • Context formatting: summary in 3–5 bullet points (example: "5 Work Products opened, 2 Work Products closed, 15+ commits, WakaTime ~32h, Day Close 4 times")

2c) Pilot RCS profile (baseline + confidence):

  • Source: Neon DB indicators.calculated_profile
  • Fields: W (worldview), M1 (self-development methods), M2 (IWE/ORZ), M3 (domain), M4 (systems), IT (infrastructure), A (agency)
  • Each slot: value (1–5) + confidence (0–100%)
  • Result: compact table with delta from the previous week (if history is available)

Output: Three data blocks ready for Step 3.


Step 3: Reflection dialog with the pilot

Run a synchronous dialog with the pilot. The goal is to convert data into insights (not just display metrics).

Questions (choose 2–3, not all):

  1. "What stands out to you in this week's numbers?" (point to 1–2 metrics from Step 2b that deviate from the trend)
  2. "What got in the way beyond what was planned?" (listen for weekly context — disruptions, urgent tasks, interface issues)
  3. "Hypothesis for the new week: what do you want to change?" (based on the answers above + RCS bottleneck from Step 2c)

Dialog format:

  • Post each question as a separate message (do not bundle multiple questions together)
  • Wait for the pilot's response
  • Based on the response, ask a follow-up or move to the next question
  • Do not ask anything after completion: the algorithm continues automatically

Output: 2–3 insights, hypothesis for the new week.


Step 4: Archive the previous weekly

Move the previous weekly from weekly/ to history/:

  1. Read DS-personal-guide/weekly/<YYYY>-Www.md (for W{N-1})
  2. Prepend a Results section to the top of the file:
    ## Results W{N-1}
    
    **Archived on:** YYYY-MM-DD  
    **Status:** closed  
    **Actual vs. hypothesis:** [from Step 3 dialog]  
    **Insights:** [brief conclusions from the dialog]  
    **Recommendation for W{N}:** [hypothesis]
  3. Move the file: weekly/<YYYY>-Www.md → history/<YYYY>-Www-results.md
  4. Commit to git: git add history/, git commit -m "Archive W{N-1}: results + hypothesis for W{N}"

Output: previous weekly archived, git committed.


Step 5: Generate hypothesis for the new week

Based on insights from Step 3, create a brief hypothesis for W{N}:

Form (fill in the blanks based on the dialog):

  • Key question: [What does the pilot want to focus on this week?]
  • Why now: [From last week's context / RCS bottleneck]
  • Key result: [One observable outcome by end of week]
  • Lever: [Which of the 7 slots (W/M1–M4/IT/A) will be strengthened this week?]

Save as a variable for Step 6.

Output: hypothesis ready.


Step 6: Write a new weekly for W{N}

Call personal_write (MCP tool) to create the file:

  • path: weekly/<YYYY>-Www.md (for W{N})
  • content: use the template from monthly-theme-template.md + hypothesis from Step 5
  • Sections:
    • Hypothesis (4–6 lines from Step 5)
    • Work Products (2–3 specific Artifacts that should appear)
    • Rhythm (slots W/M1–M4, planned time)
    • Metrics (which metrics to track for the week)
    • Comment (what to check at the end of the week)

Output: new weekly written to the DS-personal-guide repository.


Step 7: Write a daily for the first day of week W{N}

Call personal_write for the daily file:

  • path: daily/YYYY-MM-DD.md (for Monday of W{N})
  • content: scaffold with empty Day Open slots (calendar, day plan, first item from the weekly)

Output: daily created, pilot ready for Day Open on Monday.


Step 8: Report

Send one summary message (no questions):

✅ W{N-1} closed
• Weekly archived with results
• Hypothesis for W{N}: [brief phrase, max 10 words]
• New weekly ready: <YYYY>-Www.md
• Daily for Monday {date} created

→ Ready for Day Open

Output: Skill complete.


Safeguards

  • ⛔ Step 3 without dialog: If the pilot does not respond to a question, move to the next question after 20s (do not wait indefinitely)
  • ⛔ Incomplete data at Step 2: If the weekly is not found or the DB is unavailable — report this and exit the Skill (do not generate data from nothing)
  • ⛔ Double archiving: Verify the weekly has not already been archived in a previous run (check git history)

Connections

  • Entry point: protocol-close.md (week-close step 1 calls this Skill as step 0)
  • Exit point: results are ready for /day-open on Monday of W{N}
  • Tailor: The Skill calls personal_write (built-in Integration), which runs render-pilot-guides.py in the background (F28.7)
  • Memory: RCS profile is read via MCP (get_tailor_context), events from the learning DB via SQL

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.