Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when you have a spec or requirements for a multi-step task, before touching code
.claude/skills/ibrahim-3d-writing-plans/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 1572% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -20% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 5% | 0% |
Write comprehensive implementation plans assuming the engineer has zero context for our codebase and questionable taste. Document everything they need to know: which files to touch for each task, code, testing, docs they might need to check, how to test it. Give them the whole plan as bite-sized tasks. DRY. YAGNI. TDD. Frequent commits.
Assume they are a skilled developer, but know almost nothing about our toolset or problem domain. Assume they don't know good test design very well.
Announce at start: "I'm using the writing-plans skill to create the implementation plan."
Save plans to: The active conductor track directory: conductor/tracks/{track_id}/plan.md If no conductor track exists, create one first using the /orchestrator-supaconductor:new-track flow.
Each step is one action (2-5 minutes):
Every plan MUST start with this header:
markdown# [Feature Name] Implementation Plan > **For Claude:** REQUIRED SUB-SKILL: Use orchestrator-supaconductor:executing-plans to implement this plan task-by-task. **Goal:** [One sentence describing what this builds] **Architecture:** [2-3 sentences about approach] **Tech Stack:** [Key technologies/libraries] ---
`markdown### Task N: [Component Name] **Files:** - Create: `exact/path/to/file.py` - Modify: `exact/path/to/existing.py:123-145` - Test: `tests/exact/path/to/test.py` **Step 1: Write the failing test**
def test_specific_behavior(): result = function(input) assert result == expected
**Step 2: Run test to verify it fails**
Run: `pytest tests/path/test.py::test_name -v`
Expected: FAIL with "function not defined"
**Step 3: Write minimal implementation**
def function(input): return expected
**Step 4: Run test to verify it passes**
Run: `pytest tests/path/test.py::test_name -v`
Expected: PASS
**Step 5: Commit**
git add tests/path/test.py src/path/file.py git commit -m "feat: add specific feature"
This skill is part of the Conductor workflow. All plans are saved to the active track directory.
When invoked by the Conductor orchestrator (with --track-dir and --spec parameters):
--spec pathconductor/product.md, conductor/tech-stack.md{--track-dir}/plan.mdmetadata.json checkpoint to PLAN: PASSEDWhen invoked standalone (no parameters):
conductor/tracks.mdspec.md for requirementsconductor/tracks/{track_id}/plan.mdmetadata.json checkpoint to PLAN: PASSED/orchestrator-supaconductor:implement or /orchestrator-supaconductor:go to execute| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 5,082 | 7,253 | +43% | 1 | 1 | 0% | 298 | 1,248 | +319% | 0 | 0 | — |
case-02 | fail→fail | 7,250 | 1,695 | -77% | 1 | 1 | 0% | 262 | 1,129 | +331% | 0 | 0 | — |
case-03 | fail→fail | 4,357 | 2,398 | -45% | 1 | 1 | 0% | 326 | 1,165 | +257% | 0 | 0 | — |
case-04 | fail→pass | 5,051 | 14,410 | +185% | 1 | 1 | 0% | 224 | 3,746 | +1572% | 0 | 0 | — |
case-05 | pass→fail | 22,756 | 4,470 | -80% | 1 | 1 | 0% | 4,072 | 1,258 | -69% | 0 | 0 | — |
case-06 | pass→fail | 11,033 | 19,653 | +78% | 1 | 1 | 0% | 1,298 | 4,949 | +281% | 0 | 0 | — |
case-07 | fail→fail | 4,881 | 2,753 | -44% | 1 | 1 | 0% | 223 | 1,366 | +513% | 0 | 0 | — |
case-08 | fail→fail | 14,512 | 2,003 | -86% | 1 | 1 | 0% | 2,514 | 1,180 | -53% | 0 | 0 | — |
case-09 | fail→fail | 2,804 | 2,650 | -5% | 1 | 1 | 0% | 411 | 1,232 | +200% | 0 | 0 | — |
case-10 | pass→pass | 6,810 | 2,627 | -61% | 1 | 1 | 0% | 1,250 | 1,353 | +8% | 0 | 0 | — |
case-11 | fail→pass | 8,572 | 2,776 | -68% | 1 | 1 | 0% | 1,503 | 1,205 | -20% | 0 | 0 | — |
case-12 | fail→fail | 7,667 | 2,094 | -73% | 1 | 1 | 0% | 1,331 | 1,237 | -7% | 0 | 0 | — |
case-13 | fail→fail | 6,644 | 3,092 | -53% | 1 | 1 | 0% | 1,179 | 1,276 | +8% | 0 | 0 | — |
case-14 | fail→pass | 8,343 | 2,622 | -69% | 1 | 1 | 0% | 1,556 | 1,352 | -13% | 0 | 0 | — |
case-15 | fail→pass | 9,743 | 3,670 | -62% | 1 | 1 | 0% | 1,634 | 1,562 | -4% | 0 | 0 | — |
case-16 | fail→pass | 9,894 | 4,250 | -57% | 1 | 1 | 0% | 1,586 | 1,658 | +5% | 0 | 0 | — |
case-17 | fail→pass | 10,823 | 2,181 | -80% | 1 | 1 | 0% | 1,749 | 1,330 | -24% | 0 | 0 | — |
case-18 | fail→pass | 10,502 | 16,050 | +53% | 1 | 1 | 0% | 1,867 | 4,296 | +130% | 0 | 0 | — |
case-19 | fail→pass | 6,764 | 4,185 | -38% | 1 | 1 | 0% | 1,046 | 1,602 | +53% | 0 | 0 | — |
case-20 | fail→pass | 7,059 | 2,040 | -71% | 1 | 1 | 0% | 1,167 | 1,254 | +7% | 0 | 0 | — |
case-21 | pass→pass | 4,338 | 1,164 | -73% | 1 | 1 | 0% | 709 | 1,112 | +57% | 0 | 0 | — |
case-22 | fail→pass | 7,592 | 1,757 | -77% | 1 | 1 | 0% | 1,547 | 1,204 | -22% | 0 | 0 | — |
case-23 | fail→pass | 9,015 | 5,016 | -44% | 1 | 1 | 0% | 1,657 | 1,822 | +10% | 0 | 0 | — |
case-24 | fail→pass | 15,037 | 5,804 | -61% | 1 | 1 | 0% | 2,473 | 1,913 | -23% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 24 cases were attempted, and 19 counted toward the lift figure. The other 5 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +42 percentage points is the difference between those two pass rates over the 19 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.