Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Implements Manus-style file-based planning for complex tasks. Creates task_plan.md, findings.md, and progress.md. Use when starting complex multi-step tasks, research projects, or any task requiring >5 tool calls.
.claude/skills/dokhacgiakhoa-planning-with-files/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | -48% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -27% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -39% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -31% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -20% | 0% |
Work like Manus: Use persistent markdown files as your "working memory on disk."
When using this skill:
${CLAUDE_PLUGIN_ROOT}/templates/task_plan.md, findings.md, progress.md) should be created in your project directory — the folder where you're working| Location | What Goes There | |----------|-----------------| | Skill directory (${CLAUDE_PLUGIN_ROOT}/) | Templates, scripts, reference docs | | Your project directory | task_plan.md, findings.md, progress.md |
This ensures your planning files live alongside your code, not buried in the skill installation folder.
Before ANY complex task:
task_plan.md in your project — Use templates/task_plan.md as referencefindings.md in your project — Use templates/findings.md as referenceprogress.md in your project — Use templates/progress.md as reference> Note: All three planning files should be created in your current working directory (your project root), not in the skill's installation folder.
Context Window = RAM (volatile, limited)
Filesystem = Disk (persistent, unlimited)
→ Anything important gets written to disk.| File | Purpose | When to Update | |------|---------|----------------| | task_plan.md | Phases, progress, decisions | After each phase | | findings.md | Research, discoveries | After ANY discovery | | progress.md | Session log, test results | Throughout session |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 21,710 | 29,721 | +37% | 1 | 1 | 0% | 3,836 | 5,606 | +46% | 0 | 0 | — |
case-02 | fail→fail | 18,836 | 4,172 | -78% | 1 | 1 | 0% | 3,222 | 984 | -69% | 0 | 0 | — |
case-03 | fail→fail | 11,430 | 4,431 | -61% | 1 | 1 | 0% | 2,083 | 925 | -56% | 0 | 0 | — |
case-04 | fail→fail | 11,287 | 11,550 | +2% | 1 | 1 | 0% | 1,893 | 2,558 | +35% | 0 | 0 | — |
case-05 | fail→pass | 13,069 | 2,467 | -81% | 1 | 1 | 0% | 2,067 | 1,078 | -48% | 0 | 0 | — |
case-06 | fail→pass | 10,150 | 3,246 | -68% | 1 | 1 | 0% | 1,630 | 1,190 | -27% | 0 | 0 | — |
case-07 | fail→pass | 10,646 | 2,391 | -78% | 1 | 1 | 0% | 1,745 | 1,056 | -39% | 0 | 0 | — |
case-08 | fail→pass | 13,048 | 4,796 | -63% | 1 | 1 | 0% | 2,200 | 1,512 | -31% | 0 | 0 | — |
case-09 | fail→pass | 9,440 | 2,857 | -70% | 1 | 1 | 0% | 1,444 | 1,153 | -20% | 0 | 0 | — |
case-10 | fail→pass | 10,873 | 5,180 | -52% | 1 | 1 | 0% | 1,638 | 1,370 | -16% | 0 | 0 | — |
case-11 | fail→pass | 11,092 | 5,909 | -47% | 1 | 1 | 0% | 1,769 | 1,652 | -7% | 0 | 0 | — |
case-12 | fail→fail | 9,008 | 33,282 | +269% | 1 | 1 | 0% | 1,471 | 1,189 | -19% | 0 | 0 | — |
case-13 | pass→pass | 7,582 | 3,948 | -48% | 1 | 1 | 0% | 1,163 | 1,324 | +14% | 0 | 0 | — |
case-14 | fail→fail | 7,386 | 3,587 | -51% | 1 | 1 | 0% | 1,160 | 1,308 | +13% | 0 | 0 | — |
case-15 | fail→pass | 27,906 | 1,588 | -94% | 1 | 1 | 0% | 4,810 | 893 | -81% | 0 | 0 | — |
case-16 | fail→pass | 8,228 | 3,108 | -62% | 1 | 1 | 0% | 1,270 | 1,065 | -16% | 0 | 0 | — |
case-17 | pass→pass | 15,492 | 4,772 | -69% | 1 | 1 | 0% | 2,535 | 1,422 | -44% | 0 | 0 | — |
case-18 | fail→pass | 8,712 | 2,004 | -77% | 1 | 1 | 0% | 1,571 | 970 | -38% | 0 | 0 | — |
case-19 | fail→pass | 10,478 | 2,074 | -80% | 1 | 1 | 0% | 1,682 | 1,072 | -36% | 0 | 0 | — |
case-20 | pass→pass | 11,566 | 10,829 | -6% | 1 | 1 | 0% | 2,135 | 2,525 | +18% | 0 | 0 | — |
case-21 | pass→pass | 9,024 | 29,363 | +225% | 1 | 1 | 0% | 1,642 | 2,392 | +46% | 0 | 0 | — |
case-22 | pass→pass | 13,924 | 12,658 | -9% | 1 | 1 | 0% | 2,554 | 2,800 | +10% | 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. 22 cases were attempted, and 20 counted toward the lift figure. The other 2 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 +50 percentage points is the difference between those two pass rates over the 20 comparable cases.
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.