Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Maintain {{KIRO_DIR}}/steering/ as persistent project memory (bootstrap/sync). Use when initializing or updating steering documents.
.claude/skills/bilal140202-kiro-steering/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | -7% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 44% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 17% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 22% | 0% |
You are a specialized skill for maintaining {{KIRO_DIR}}/steering/ as persistent project memory.
Role: Maintain {{KIRO_DIR}}/steering/ as persistent project memory.
Mission:
Success Criteria:
{{KIRO_DIR}}/steering/*.md treated equally (core + custom)If steering context is already available from conversation, skip redundant file reads.
{{KIRO_DIR}}/settings/templates/steering/{{KIRO_DIR}}/steering/*.md filesrules/steering-principles.md from this skill's directory for steering principlesCheck {{KIRO_DIR}}/steering/ status:
Bootstrap Mode: Empty OR missing core files (product.md, tech.md, structure.md) Sync Mode: All core files exist
{{KIRO_DIR}}/settings/templates/steering/The following research areas are independent and can be executed in parallel:
After all parallel research completes, synthesize patterns for steering files.
rules/steering-principles.md from this skill's directoryFocus: Patterns that guide decisions, not catalogs of files/dependencies.
{{KIRO_DIR}}/steering/*.md)Update Philosophy: Add, don't replace. Preserve user sections.
From rules/steering-principles.md (in this skill's directory):
> "If new code follows existing patterns, steering shouldn't need updating."
Document patterns and principles, not exhaustive lists.
Bad: List every file in directory tree Good: Describe organization pattern with examples
Glob: Find source/config filesRead: Read steering, docs, configsGrep: Search patternsBash with ls: Analyze structureJIT Strategy: Fetch when needed, not upfront.
Chat summary only (files updated directly).
Steering Created
## Generated:
- product.md: [Brief description]
- tech.md: [Key stack]
- structure.md: [Organization]
Review and approve as Source of Truth.Steering Updated
## Changes:
- tech.md: React 18 → 19
- structure.md: Added API pattern
## Code Drift:
- Components not following import conventions
## Recommendations:
- Consider api-standards.mdInput: Empty steering, React TypeScript project Output: 3 files with patterns - "Feature-first", "TypeScript strict", "React 19"
Input: Existing steering, new /api directory Output: Updated structure.md, flagged non-compliant files, suggested api-standards.md
{{KIRO_DIR}}/steering/*.md loaded as project memory{{KIRO_DIR}}/settings/ content should NOT be documented in steering files (settings are metadata, not project knowledge)| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-21 | fail→fail | 2,741 | 5,157 | +88% | 1 | 1 | 0% | 213 | 1,249 | +486% | 0 | 0 | — |
case-01 | fail→fail | 4,642 | 4,829 | +4% | 1 | 1 | 0% | 263 | 1,332 | +406% | 0 | 0 | — |
case-02 | fail→fail | 10,540 | 5,653 | -46% | 1 | 1 | 0% | 1,679 | 1,361 | -19% | 0 | 0 | — |
case-03 | fail→fail | 7,121 | 7,924 | +11% | 1 | 1 | 0% | 1,315 | 1,294 | -2% | 0 | 0 | — |
case-04 | fail→fail | 13,418 | 4,387 | -67% | 1 | 1 | 0% | 2,360 | 1,661 | -30% | 0 | 0 | — |
case-05 | fail→pass | 10,644 | 3,570 | -66% | 1 | 1 | 0% | 1,759 | 1,639 | -7% | 0 | 0 | — |
case-06 | fail→fail | 2,859 | 7,823 | +174% | 1 | 1 | 0% | 340 | 1,411 | +315% | 0 | 0 | — |
case-07 | pass→pass | 11,739 | 6,008 | -49% | 1 | 1 | 0% | 1,969 | 2,034 | +3% | 0 | 0 | — |
case-08 | fail→pass | 8,257 | 4,958 | -40% | 1 | 1 | 0% | 1,322 | 1,910 | +44% | 0 | 0 | — |
case-09 | fail→pass | 9,923 | 4,567 | -54% | 1 | 1 | 0% | 1,552 | 1,810 | +17% | 0 | 0 | — |
case-10 | pass→pass | 10,269 | 4,706 | -54% | 1 | 1 | 0% | 1,732 | 1,899 | +10% | 0 | 0 | — |
case-11 | pass→pass | 5,685 | 2,593 | -54% | 1 | 1 | 0% | 944 | 1,544 | +64% | 0 | 0 | — |
case-12 | fail→pass | 8,857 | 1,590 | -82% | 1 | 1 | 0% | 1,457 | 1,313 | -10% | 0 | 0 | — |
case-13 | fail→pass | 9,527 | 4,752 | -50% | 1 | 1 | 0% | 1,478 | 1,803 | +22% | 0 | 0 | — |
case-14 | fail→pass | 8,524 | 5,026 | -41% | 1 | 1 | 0% | 1,401 | 1,893 | +35% | 0 | 0 | — |
case-15 | pass→pass | 8,614 | 3,174 | -63% | 1 | 1 | 0% | 1,334 | 1,544 | +16% | 0 | 0 | — |
case-16 | fail→fail | 7,013 | 3,000 | -57% | 1 | 1 | 0% | 1,145 | 1,598 | +40% | 0 | 0 | — |
case-17 | fail→pass | 5,294 | 5,146 | -3% | 1 | 1 | 0% | 774 | 1,887 | +144% | 0 | 0 | — |
case-18 | fail→pass | 12,501 | 6,664 | -47% | 1 | 1 | 0% | 2,097 | 2,083 | -1% | 0 | 0 | — |
case-19 | fail→pass | 11,331 | 5,521 | -51% | 1 | 1 | 0% | 1,788 | 1,888 | +6% | 0 | 0 | — |
case-20 | pass→fail | 2,826 | 6,289 | +123% | 1 | 1 | 0% | 282 | 1,378 | +389% | 0 | 0 | — |
case-22 | pass→fail | 14,734 | 5,559 | -62% | 1 | 1 | 0% | 2,771 | 1,321 | -52% | 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 15 counted toward the lift figure. The other 7 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 +32 percentage points is the difference between those two pass rates over the 15 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.