UNPKG

arela

Version:

AI-powered CTO with multi-agent orchestration, code summarization, visual testing (web + mobile) for blazing fast development.

244 lines (184 loc) 4.75 kB
# [AGENT]-[ID]: [Title] **Agent:** [claude | codex | cascade] **Priority:** [critical | high | medium | low] **Complexity:** [simple | medium | complex] **Estimated Time:** [hours] --- ## 🔍 BEFORE YOU START (Read This First!) **Use Arela's Hexi-Memory System to save 85k+ tokens:** ### 1. Semantic Search (FASTEST - Use This First!) ```bash arela_search "slice extraction logic" arela_search "import path rewriting" arela_search "test runner implementation" ``` - **Cost:** ~1k tokens per search - **Finds:** Relevant code semantically (Vector Memory) - **When:** Always try this FIRST ### 2. Check Project Memory ```bash # Get project patterns & conventions arela memory project --key auth_patterns arela memory project --key testing_strategy arela memory project --category decision # What's in Project Memory: # - Architecture decisions & rationales # - Patterns & conventions (observed, not claimed) # - Project-scoped todos & tech debt # - Component summaries ``` - **Cost:** 0 tokens (local SQLite) - **Finds:** "How we do things in THIS repo" ### 3. Check User Memory (Cross-Project Patterns) ```bash # Get your preferences across all projects arela memory user --key preferred_db arela memory user --key testing_style arela memory user --patterns # What's in User Memory: # - Your tech stack preferences # - Workflow preferences (PR size, TDD, etc.) # - Expertise levels # - Patterns you follow (or avoid) ``` - **Cost:** 0 tokens (global SQLite) - **Finds:** "How YOU typically work" ### 4. Check Architecture Graph ```bash # See dependencies & relationships arela graph --from src/auth/login.ts arela graph --slice auth # What's in Graph Memory: # - File → file imports # - Symbol → symbol calls # - Slices / module boundaries ``` - **Cost:** 0 tokens (local SQLite) - **Finds:** Structural relationships ### 5. Check Governance Log ```bash # See past decisions & why arela memory governance --recent 10 arela memory governance --search "authentication" # What's in Governance Memory: # - Architectural decisions # - Tooling choices # - Policy/governance rules # - Linked artifacts (docs, PRs, ADRs) ``` - **Cost:** 0 tokens (append-only log) - **Finds:** "What we decided and why" ### 6. grep/find (LAST RESORT) - Only use if memory system + semantic search fails - **Cost:** 85k+ tokens for large searches - Use sparingly! --- ## Context **Why this task exists:** [Explain the problem or need] **Current state:** [What exists now] **Desired state:** [What we want after this ticket] --- ## Requirements ### Must Have - [ ] Requirement 1 - [ ] Requirement 2 - [ ] Requirement 3 ### Should Have - [ ] Nice-to-have 1 - [ ] Nice-to-have 2 ### Nice to Have - [ ] Future enhancement 1 - [ ] Future enhancement 2 --- ## Technical Implementation ### Files to Create/Modify ``` src/ ├── feature/ ├── index.ts ├── implementation.ts └── types.ts ``` ### Architecture ```typescript // High-level design interface Example { // Key interfaces } async function mainFlow() { // 1. Step one // 2. Step two // 3. Step three } ``` ### Key Decisions - **Decision 1:** [Why we chose this approach] - **Decision 2:** [Trade-offs considered] --- ## Acceptance Criteria - [ ] Feature works as specified - [ ] Tests pass (unit + integration) - [ ] Documentation updated - [ ] No breaking changes (or documented) - [ ] Code reviewed - [ ] Performance acceptable --- ## Test Plan ### Unit Tests - Test case 1 - Test case 2 ### Integration Tests - Integration scenario 1 - Integration scenario 2 ### Edge Cases - Edge case 1 - Edge case 2 --- ## Success Metrics **Performance:** - Metric 1: < X seconds - Metric 2: < Y operations **Quality:** - Test coverage: > 80% - No regressions - Clean build --- ## Example Output ```bash $ arela command --option 🚀 Starting... Step 1 complete Step 2 complete Done! Summary: - Item 1: Success - Item 2: Success ``` --- ## Documentation **Files to create/update:** - `docs/feature.md` - Complete guide - `README.md` - Add feature to list - `CHANGELOG.md` - Document changes --- ## Notes **Important considerations:** - Note 1 - Note 2 **Related work:** - Ticket #123 - ADR-042 --- ## Remember **Hexi-Memory Search Priority:** 1. `arela_search` - Semantic code search (1k tokens) 2. `arela memory project` - This repo's patterns (0 tokens) 3. `arela memory user` - Your preferences (0 tokens) 4. `arela graph` - Dependencies & structure (0 tokens) 5. `arela memory governance` - Past decisions (0 tokens) 6. `grep`/`find` - Only if memory fails (85k+ tokens) **The memory system knows more than you think. Use it!** 🧠 **Save tokens, save money, ship faster!** 🚀