agentleFS
chrisblattmanchrisblattman/claudeblattman/skills/goals-review/SKILL.md

goals-review

Skill479 starsChanged 23 days agoMITView on GitHub

Review quarterly objectives, progress scores and deadlines using the configured goals file.

The file

We split this skill into the rules it gives an agent

12 rules. About 0 are common, 9 are distinctive, and 3 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.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    All objectives with key results and current progress scores

    line 69
  2. Very specific:

    If goals.yaml is missing, prompt the user to create it from the template.

    line 74
  3. Distinctive:

    Search Granola for meetings in last 2 weeks mentioning papers, manuscripts, or writing

    line 83
  4. Distinctive:

    Check Gmail for recent co-author emails (sent items) to gauge feedback turnaround

    line 84
  5. Distinctive:

    Check Reminders (via osascript, macOS) for grant-related items

    line 87
  6. Distinctive:

    Search Gmail for "grant report" or "progress report" in last 30 days

    line 88
  7. Distinctive:

    Check Gmail sent folder for team member response patterns

    line 91
  8. Distinctive:

    Search Granola for recent project meetings

    line 92
  9. Distinctive:

    If any source is unavailable, note it and continue — never block on a failed query

    line 96
  10. Very specific:

    After the user provides updates, write the updated goals.yaml file.

    line 149
  11. Very specific:

    Update next_review in goals.yaml to 14 days from today.

    line 153
  12. Distinctive:

    Check if a biweekly reminder exists.

    line 155

≈ “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

Goals Review

v1.1 — Adapted for public use. Review quarterly objectives, update progress scores, and manage deadlines.

Review quarterly objectives, update progress scores, surface deadlines, and recalibrate priorities. Run biweekly or on demand.

Prerequisites

Required:

  • ~/.claude-assistant/config/goals.yaml — objectives file (see First-Time Setup)

Optional (for evidence gathering):

  • Gmail MCP — for email activity evidence
  • Google Calendar MCP — for meeting alignment evidence
  • Granola MCP — for meeting transcript evidence
  • Apple Reminders (macOS only) — for task-related evidence

First-Time Setup

  1. Create config directory:
    mkdir -p ~/.claude-assistant/config
    mkdir -p ~/.claude-assistant/logs
  2. Download goals template:
    curl -o ~/.claude-assistant/config/goals.yaml \
      https://raw.githubusercontent.com/chrisblattman/claudeblattman/main/templates/goals-yaml-template.yaml
  3. Edit goals.yaml with your objectives, weights, and key results.
  4. Test with status view:
    /goals-review status

Customization Points

SettingWhere to ConfigureDefault
ModelFrontmatter model: linesonnet (remove to use default)
Objectivesgoals.yaml → objectivesTemplate examples
Push levelgoals.yaml → meta.push_levelmoderate
Review frequencyStep 5 belowBiweekly (14 days)
Evidence sourcesStep 2 belowAll available MCPs
Performance logStep 6 below~/.claude-assistant/logs/skill-performance.csv

Arguments

$ARGUMENTS can include:

  • (none) — full review with interactive progress updates
  • status — quick dashboard only (no interactive updates)
  • deadlines — show only upcoming deadlines

Instructions

Step 1: Read Goals Config

Read ~/.claude-assistant/config/goals.yaml.

Extract:

  • All objectives with key results and current progress scores
  • push_level setting
  • upcoming_deadlines list
  • next_review date

If goals.yaml is missing, prompt the user to create it from the template.

Step 2: Gather Evidence (parallel where possible)

For each active objective, gather recent signals using available MCP tools:

Example patterns by objective type:

Research Output:

  • Search Granola for meetings in last 2 weeks mentioning papers, manuscripts, or writing
  • Check Gmail for recent co-author emails (sent items) to gauge feedback turnaround

Grant Management:

  • Check Reminders (via osascript, macOS) for grant-related items
  • Search Gmail for "grant report" or "progress report" in last 30 days

Team Effectiveness:

  • Check Gmail sent folder for team member response patterns
  • Search Granola for recent project meetings

General pattern:

  • Match each objective's name and key_results against email subjects, meeting titles, and reminder names
  • If any source is unavailable, note it and continue — never block on a failed query
  • Cap at 3 searches per objective

Step 3: Compute Progress Dashboard

For each objective, determine status:

  • On track — progress scores advancing, no overdue key results
  • At risk — some key results stalled or behind expected pace
  • Behind — multiple key results with no progress, approaching deadlines missed

Display:

# Goals Review — [Date]
Quarter: [quarter] | Push level: [level] | Next review: [date]

────────────────────

## 1. [Objective name] (weight: [N]%)
   Status: [ON TRACK / AT RISK / BEHIND]

   Key Results:
   a) [description] ............ [progress] → [suggested update]
      Evidence: [1-line summary of what was found]
   b) [description] ............ [progress] → [suggested update]
      Evidence: [1-line summary]

────────────────────

## 2. [Next objective...]
   ...

────────────────────

## Upcoming Deadlines (next 30 days)
- [date]: [description] (objective: [name])
- ...
[If none: "No deadlines in the next 30 days"]

## Stalled Areas
- [Any key result with 0.0 progress for 2+ review cycles]

## Recommendation
[1-2 sentences on what to focus on this fortnight]

Step 4: Interactive Update (skip if status or deadlines argument)

For each objective, ask:

  • "Update progress for [key result]? Current: [score]. Enter new score (0.0-1.0) or press Enter to keep."
  • "Any deadlines to add for the next 30 days?"
  • "Should push_level change? Current: [level]"

After the user provides updates, write the updated goals.yaml file.

Step 5: Set Next Review

Update next_review in goals.yaml to 14 days from today.

Check if a biweekly reminder exists. If not, suggest: "Create a recurring biweekly reminder 'Run /goals-review with Claude'?"

Step 6: Log Performance

echo "$(date +%Y-%m-%d),goals-review,TOOL_CALLS,NOTES" >> ~/.claude-assistant/logs/skill-performance.csv

Replace TOOL_CALLS with approximate count and NOTES with brief summary.

Error Handling

  • Granola unavailable: Skip meeting evidence, note "Granola unavailable"
  • Gmail unavailable: Skip email evidence, note "Gmail unavailable"
  • osascript fails: Skip reminder evidence
  • goals.yaml missing: Prompt user to create from template

Integration Notes

  • Manages the goals.yaml file that /checkin and /morning-brief read for goal alignment. Keep it updated here, and those skills automatically benefit.
  • Run biweekly to keep progress scores current. Set up a recurring reminder.
  • Push level affects how aggressively /checkin and /morning-brief nudge about calendar alignment and deep work time.

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.