flutter-lessons-patterns
Skill1 starsChanged 40 days agoApache-2.0View on GitHub
Cross-project Flutter patterns distilled from CarSah + Hermex_Android + Azdal — 54 programming patterns, each classified GATE/RULE/JUDGMENT + RFC 2119 negation (MUST/SHOULD/MAY), plus missing-gate detection lenses. Single source of truth for all Flutter/Dart/Android coding lessons. Load before every implementation task. Full bodies live in references/ (split 2026-08-22 for size).
In this repository18 other files it gives its agents.
- agent-repo-handoff-loopSkill
- device-screen-verificationSkill
- flutter-android-build-systemSkill
- flutter-design-anti-patternsSkill
- flutter-ios-build-systemSkill
- flutter-isar-clean-arch-setupSkill
- flutter-isar-testingSkill
- flutter-patternsSkill
- flutter-sdk-changelogSkill
- flutter-soul-stewardshipSkill
- github-project-auditSkill
- mcp-bridge-patternsSkill
- repo-front-doorSkill
- skill-ecosystem-syncSkill
- specification-writingSkill
- supabase-fullstackSkill
- translation-fidelitySkill
- what-if-analysisSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 37
Every pattern is classified by two exclusion filters then a weighted score:
- Distinctive: line 49
When the referenced file grows past ~50K, split again — one file per era, same convention.
≈ “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.
Files
Flutter Cross-Project Patterns — Unified Programming Lessons
Sources: CarSah (LL-001–LL-020) + Hermex_Android (LL-001–LL-042) + Azdal Patterns: 54 — full bodies are in the four reference files below (split 2026-08-22 because SKILL.md exceeded 100K and patches were being rejected). Purpose: SINGLE source of truth for ALL Flutter/Dart/Android programming lessons. Load before every implementation task. Governance/process lessons belong in
agent-repo-handoff-loop(referenced from there) — this skill is code-level.
Enforcement levels (adopted 2026-08-11 — founder decision)
Every pattern is classified by two exclusion filters then a weighted score:
- 🚪 GATE — machine-enforced: a lint/test/script fails the build if violated. Universal truth + costly to retrofit.
- 📏 RULE — documented convention enforced by review. Universal truth but not machine-detectable.
- 🧭 JUDGMENT — project decision: the skill gives criteria, the answer depends on project context.
Negation axis (RFC 2119 mapping): 🚪 MUST / MUST NOT · 📏 SHOULD / SHOULD NOT · 🧭 MAY.
How to read this skill (IMPORTANT — this file is now an index)
- Read the pattern you need from the reference file that carries it (map below).
- Always cite patterns by number in reviews, letters, and code comments (Pattern 52, etc.) — the shared vocabulary.
- New patterns (55+) must be added directly to the appropriate references/ file as
## Pattern N — ...blocks — NEVER appended to SKILL.md (it will overflow again). - When the referenced file grows past ~50K, split again — one file per era, same convention.
Pattern Index (54 patterns)
| # | Pattern | File |
| 1 | Provider Invalidation Rule (LL-003) | patterns-01-18-core-android-mobile |
| 2 | Device Verification Gate (LL-002 + LL-013) | patterns-01-18-core-android-mobile |
| 3 | Zero Hardcoded Strings (LL-001 + LL-006 + LL-018) | patterns-01-18-core-android-mobile |
| 4 | Save-Gating Validators (LL-005 + LL-017) | patterns-01-18-core-android-mobile |
| 5 | Tests in Same PR (LL-007) | patterns-01-18-core-android-mobile |
| 6 | Spec Sync Gate (LL-008) | patterns-01-18-core-android-mobile |
| 7 | Design Before Implementation (LL-016 + LL-019) | patterns-01-18-core-android-mobile |
| 8 | 1-Day BL Maximum (LL-020) | patterns-01-18-core-android-mobile |
| 9 | Android Namespace = MainActivity Package (LL-024) | patterns-01-18-core-android-mobile |
| 10 | Isar + ProGuard = Crash (LL-025) | patterns-01-18-core-android-mobile |
| 11 | Official Android Sources Mandatory (LL-026) | patterns-01-18-core-android-mobile |
| 12 | Design Quality Anti-Patterns (LL-027 — Cross-Reference) | patterns-01-18-core-android-mobile |
| 13 | State Mutation Order: Snapshot BEFORE Mutating (LL-029) | patterns-01-18-core-android-mobile |
| 14 | Silent API Key Redaction: *** Literal (LL-022) | patterns-01-18-core-android-mobile |
| 15 | Fake Connection State: Never Set 'connected' Without Health Check (LL-023) | patterns-01-18-core-android-mobile |
| 16 | API Query Parameters: Always Pass include_disabled=true (LL-042) | patterns-01-18-core-android-mobile |
| 17 | Verify On Disk Before Claiming (Meta-Pattern — Sulaiman Session 2026-07-11) | patterns-01-18-core-android-mobile |
| 18 | ProviderScope in Widget Tests: Mirror main.dart Exactly (LL-018) | patterns-01-18-core-android-mobile |
| 19 | Empty Catch Blocks FORBIDDEN in Security Paths (LL-019) | patterns-19-30-extended-meta |
| 20 | GoRouter: Static Routes BEFORE Parameterized (LL-005) | patterns-19-30-extended-meta |
| 21 | Provider Hygiene: NotifierProvider vs NotifierProvider.autoDispose (LL-003) | patterns-19-30-extended-meta |
| 22 | Repository Null-Safety: Accept Nullable Dependencies (LL-006) | patterns-19-30-extended-meta |
| 23 | isBusy Guard: Provider-Level Atomicity (LL-004) | patterns-19-30-extended-meta |
| 24 | Router Wiring Gate: Feature NOT Done Until Wired (LL-017) | patterns-19-30-extended-meta |
| 25 | Provider Invalidation: Widget Layer, Not Provider Internals (LL-007) | patterns-19-30-extended-meta |
| 26 | SSE Parser: Custom Format Handling (LL-002 + HERMEX-007) | patterns-19-30-extended-meta |
| 27 | Branch Hygiene: Verify Baseline Before Starting Work (LL-033) | patterns-19-30-extended-meta |
| 28 | [DUPLICATE — See Pattern 13] SSE Duplicate Message Prevention: Snapshot BEFORE Streaming (LL-029 — Extended) | patterns-19-30-extended-meta |
| 29 | API Response Key Format: Use data, Not messages (LL-041) | patterns-19-30-extended-meta |
| 30 | Build Responsibility: Lead Architect Coordinates, DevOps Builds (LL-044) | patterns-19-30-extended-meta |
| 31 | Lessons Flow to Shared Knowledge Base (LL-045) | patterns-31-44-agents-widgets-rc6 |
| 32 | Impact Analysis Before Implementation (LL-046) | patterns-31-44-agents-widgets-rc6 |
| 33 | Stored First Decision: Never Re-Call Non-Deterministic APIs | patterns-31-44-agents-widgets-rc6 |
| 34 | Riverpod Reactive Service: Bridge Platform SDK Callbacks to StateNotifier | patterns-31-44-agents-widgets-rc6 |
| 35 | Compute Derived Values Locally: Never Trust LLM Math | patterns-31-44-agents-widgets-rc6 |
| 36 | Ephemeral Message Lifecycle: Track Id, Remove, Replace | patterns-31-44-agents-widgets-rc6 |
| 37 | LLM Must Not Emit Actionable UI: App Constructs UI From Verified Data | patterns-31-44-agents-widgets-rc6 |
| 38 | Android INTERNET Permission on Custom OEM ROMs | patterns-31-44-agents-widgets-rc6 |
| 39 | Widget "Answered Once": Buttons Disabled After First Action | patterns-31-44-agents-widgets-rc6 |
| 40 | Full-File Rewrite Callback Verification | patterns-31-44-agents-widgets-rc6 |
| 41 | Error-Handling Architecture: validateStatus + interceptor + sanitizeError (LL-047) | patterns-31-44-agents-widgets-rc6 |
| 42 | Certificate Pinning Uniform Wiring: Single ApiClient Provider (LL-048) | patterns-31-44-agents-widgets-rc6 |
| 43 | Reactive Profile Switching: Watch connectionProvider (LL-049) | patterns-31-44-agents-widgets-rc6 |
| 44 | Gate Rescan Integrity: Re-test SPECIFIC Rejected Findings (LL-050 / ADR-012) | patterns-31-44-agents-widgets-rc6 |
| 45 | Bundled-Task Pattern for Shared-File Conflicts (LL-051) | patterns-45-54-stage4-design-meta |
| 46 | Never Say "No Data": Cold Start Intelligence (LL-009) | patterns-45-54-stage4-design-meta |
| 47 | Live-Device Verification Supremacy (LL-010) | patterns-45-54-stage4-design-meta |
| 48 | Regex Pre-Filter Gates + Disabled Button Colors (LL-011) | patterns-45-54-stage4-design-meta |
| 49 | Design System Architecture: Tokens Gate, Sub-Themes Rule, Components Judgment (LL-052) | patterns-45-54-stage4-design-meta |
| 50 | Missing-Gate Detection: Run Lenses, Not Memory (LL-053) | patterns-45-54-stage4-design-meta |
| 51 | go_router Shells: relative child paths + measured-resolved-behaviour (LL-054) | patterns-45-54-stage4-design-meta |
| 52 | Repeated CI Failures Are a Forensic Investigation, Not a Push Race (LL-055) | patterns-45-54-stage4-design-meta |
| 53 | One Attempt Is Half the Truth: Open Every Attempt Before Claiming Attribution (LL-056) | patterns-45-54-stage4-design-meta |
| 54 | The Linter's Suggestion Is Not Always the Fix: Conflicting Analyzers (LL-057) | patterns-45-54-stage4-design-meta |
| 55 | Process & Governance Patterns: Phased QA, EPIC Gates, Evidence Discipline (multi-project post-mortems) | process-governance-patterns |
Note on numbering: the reference files keep their original
## Pattern Nheadings (the extraction was mechanical — content is identical to v2.24). The index above is for navigation; the real mapping is Pattern N → reference file, not the index row numbers.
Skill divided 2026-08-22 (founder decision). Original v2.24 content preserved byte-for-byte inside the four reference files. Version 2.25 — index-only SKILL.md, full content in references/.
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
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.