clawpowers 1.1.4 → 2.0.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (75) hide show
  1. package/CHANGELOG.md +94 -0
  2. package/LICENSE +44 -0
  3. package/README.md +204 -228
  4. package/SECURITY.md +72 -0
  5. package/dist/index.d.ts +844 -0
  6. package/dist/index.js +2536 -0
  7. package/dist/index.js.map +1 -0
  8. package/package.json +50 -44
  9. package/.claude-plugin/manifest.json +0 -19
  10. package/.codex/INSTALL.md +0 -36
  11. package/.cursor-plugin/manifest.json +0 -21
  12. package/.opencode/INSTALL.md +0 -52
  13. package/ARCHITECTURE.md +0 -69
  14. package/bin/clawpowers.js +0 -625
  15. package/bin/clawpowers.sh +0 -91
  16. package/docs/demo/clawpowers-demo.cast +0 -197
  17. package/docs/demo/clawpowers-demo.gif +0 -0
  18. package/docs/launch-images/25-skills-breakdown.jpg +0 -0
  19. package/docs/launch-images/clawpowers-vs-superpowers.jpg +0 -0
  20. package/docs/launch-images/economic-code-optimization.jpg +0 -0
  21. package/docs/launch-images/native-vs-bridge-2.jpg +0 -0
  22. package/docs/launch-images/native-vs-bridge.jpg +0 -0
  23. package/docs/launch-images/post1-hero-lobster.jpg +0 -0
  24. package/docs/launch-images/post2-dashboard.jpg +0 -0
  25. package/docs/launch-images/post3-superpowers.jpg +0 -0
  26. package/docs/launch-images/post4-before-after.jpg +0 -0
  27. package/docs/launch-images/post5-install-now.jpg +0 -0
  28. package/docs/launch-images/ultimate-stack.jpg +0 -0
  29. package/docs/launch-posts.md +0 -76
  30. package/docs/quickstart-first-transaction.md +0 -204
  31. package/gemini-extension.json +0 -32
  32. package/hooks/session-start +0 -205
  33. package/hooks/session-start.cmd +0 -43
  34. package/hooks/session-start.js +0 -163
  35. package/runtime/demo/README.md +0 -78
  36. package/runtime/demo/x402-mock-server.js +0 -230
  37. package/runtime/feedback/analyze.js +0 -621
  38. package/runtime/feedback/analyze.sh +0 -546
  39. package/runtime/init.js +0 -210
  40. package/runtime/init.sh +0 -178
  41. package/runtime/metrics/collector.js +0 -361
  42. package/runtime/metrics/collector.sh +0 -308
  43. package/runtime/payments/ledger.js +0 -305
  44. package/runtime/payments/ledger.sh +0 -262
  45. package/runtime/payments/pipeline.js +0 -455
  46. package/runtime/persistence/store.js +0 -433
  47. package/runtime/persistence/store.sh +0 -303
  48. package/skill.json +0 -106
  49. package/skills/agent-bounties/SKILL.md +0 -553
  50. package/skills/agent-payments/SKILL.md +0 -479
  51. package/skills/brainstorming/SKILL.md +0 -233
  52. package/skills/content-pipeline/SKILL.md +0 -282
  53. package/skills/cross-project-knowledge/SKILL.md +0 -345
  54. package/skills/dispatching-parallel-agents/SKILL.md +0 -305
  55. package/skills/economic-code-optimization/SKILL.md +0 -265
  56. package/skills/executing-plans/SKILL.md +0 -255
  57. package/skills/finishing-a-development-branch/SKILL.md +0 -260
  58. package/skills/formal-verification-lite/SKILL.md +0 -441
  59. package/skills/learn-how-to-learn/SKILL.md +0 -235
  60. package/skills/market-intelligence/SKILL.md +0 -323
  61. package/skills/meta-skill-evolution/SKILL.md +0 -325
  62. package/skills/prospecting/SKILL.md +0 -454
  63. package/skills/receiving-code-review/SKILL.md +0 -225
  64. package/skills/requesting-code-review/SKILL.md +0 -206
  65. package/skills/security-audit/SKILL.md +0 -353
  66. package/skills/self-healing-code/SKILL.md +0 -369
  67. package/skills/subagent-driven-development/SKILL.md +0 -244
  68. package/skills/systematic-debugging/SKILL.md +0 -355
  69. package/skills/test-driven-development/SKILL.md +0 -416
  70. package/skills/using-clawpowers/SKILL.md +0 -160
  71. package/skills/using-git-worktrees/SKILL.md +0 -261
  72. package/skills/validator/SKILL.md +0 -281
  73. package/skills/verification-before-completion/SKILL.md +0 -254
  74. package/skills/writing-plans/SKILL.md +0 -276
  75. package/skills/writing-skills/SKILL.md +0 -260
@@ -1,261 +0,0 @@
1
- ---
2
- name: using-git-worktrees
3
- description: Manage isolated Git worktrees for parallel branch development. Activate when you need to work on multiple branches simultaneously or isolate subagent work.
4
- version: 1.0.0
5
- requires:
6
- tools: [git, bash]
7
- runtime: false
8
- metrics:
9
- tracks: [worktrees_created, conflicts_encountered, isolation_violations, lifecycle_completion_rate]
10
- improves: [conflict_prediction, worktree_naming, cleanup_timing]
11
- ---
12
-
13
- # Using Git Worktrees
14
-
15
- ## When to Use
16
-
17
- Apply this skill when:
18
-
19
- - Working on 2+ branches simultaneously without switching
20
- - Running subagents in parallel (each needs its own working directory)
21
- - Testing a feature while bug-fixing on another branch
22
- - Reviewing a colleague's branch while continuing your own work
23
- - Running long-running processes (tests, builds) on one branch while editing another
24
-
25
- **Skip when:**
26
- - You only have one branch active at a time
27
- - Your editor doesn't handle multiple root directories well
28
- - The branches share files that would conflict on disk (same path, different content)
29
-
30
- ## Core Methodology
31
-
32
- ### Understanding Worktrees
33
-
34
- A Git worktree is a separate working directory linked to the same repository. Each worktree:
35
- - Has its own checked-out branch
36
- - Has its own working tree state (staged/unstaged changes)
37
- - Shares the repository's history, objects, and refs
38
- - Cannot have the same branch checked out as another worktree
39
-
40
- ```
41
- .git/ ← Shared repository database
42
- worktrees/
43
- feature-auth/ ← Worktree metadata
44
- feature-payments/ ← Worktree metadata
45
-
46
- ../feature-auth/ ← Separate directory on disk
47
- src/
48
- tests/
49
-
50
- ../feature-payments/ ← Separate directory on disk
51
- src/
52
- tests/
53
- ```
54
-
55
- ### Worktree Lifecycle
56
-
57
- #### Create
58
-
59
- ```bash
60
- # Create worktree for existing branch
61
- git worktree add ../feature-auth feature/auth-service
62
-
63
- # Create worktree and new branch simultaneously
64
- git worktree add -b feature/payments ../feature-payments main
65
-
66
- # Create worktree from specific commit
67
- git worktree add ../hotfix-3.1 v3.1.0
68
- ```
69
-
70
- **Naming convention for parallel subagent work:**
71
- ```bash
72
- # Use task or feature name as both branch and directory
73
- git worktree add ../clawpowers-task-auth feature/task-auth
74
- git worktree add ../clawpowers-task-db feature/task-db
75
- git worktree add ../clawpowers-task-api feature/task-api
76
- ```
77
-
78
- #### Verify
79
-
80
- ```bash
81
- git worktree list
82
- # output:
83
- # /Users/you/project a3f9b2c [main]
84
- # /Users/you/feature-auth 0000000 [feature/auth-service]
85
- # /Users/you/feature-payments 0000000 [feature/payments]
86
- ```
87
-
88
- #### Work in the Worktree
89
-
90
- Each worktree is a full working directory. Navigate to it and work normally:
91
-
92
- ```bash
93
- cd ../feature-auth
94
- git status # Independent of main working tree
95
- git add src/auth.py
96
- git commit -m "feat(auth): implement JWT issuance"
97
- ```
98
-
99
- Changes in one worktree are invisible to others until merged.
100
-
101
- #### Sync with Main
102
-
103
- When you need to update a worktree with latest main:
104
-
105
- ```bash
106
- cd ../feature-auth
107
- git fetch origin
108
- git rebase origin/main # Preferred: linear history
109
- # or
110
- git merge origin/main # If rebase would cause conflicts
111
- ```
112
-
113
- Run `git worktree list` first — if another worktree has the same base, check for merge conflicts proactively.
114
-
115
- #### Cleanup
116
-
117
- When the branch is merged:
118
-
119
- ```bash
120
- # From main repository directory
121
- git worktree remove ../feature-auth # Removes directory
122
- git branch -d feature/auth-service # Remove branch
123
-
124
- # If the worktree has uncommitted changes and you want to force:
125
- git worktree remove --force ../feature-auth
126
-
127
- # List remaining worktrees to verify
128
- git worktree list
129
- ```
130
-
131
- **Cleanup checklist:**
132
- - [ ] Branch is merged to main (or PR is approved)
133
- - [ ] Worktree has no uncommitted changes
134
- - [ ] No processes are running in the worktree directory
135
- - [ ] Remove directory, then remove branch
136
-
137
- ### Conflict Prevention
138
-
139
- Worktrees share the index but have separate working trees. Common conflicts:
140
-
141
- **Same branch in two worktrees:** Git prevents this — you'll get an error:
142
- ```
143
- fatal: 'feature/auth-service' is already checked out
144
- ```
145
-
146
- **Solution:** Use separate branches even for related work.
147
-
148
- **Both worktrees editing the same file:** Legal, but merging will require conflict resolution:
149
- ```bash
150
- # Check overlap before creating worktrees
151
- git diff --name-only main..feature/branch-a
152
- git diff --name-only main..feature/branch-b
153
- # If outputs overlap, consider sequential rather than parallel work
154
- ```
155
-
156
- **Submodule issues:** Worktrees and submodules interact poorly. If your repo uses submodules, test worktree creation in a non-submodule path first.
157
-
158
- ### Pattern: Subagent Work Isolation
159
-
160
- The primary ClawPowers use case: give each subagent its own worktree.
161
-
162
- ```bash
163
- # Main orchestrator creates worktrees
164
- TASKS=("auth" "db" "api" "tests")
165
- for task in "${TASKS[@]}"; do
166
- git worktree add "../${REPO_NAME}-task-${task}" -b "feature/task-${task}" main
167
- echo "Created worktree for task-${task} at ../${REPO_NAME}-task-${task}"
168
- done
169
-
170
- # Each subagent receives its worktree path
171
- # Subagent-auth works in: ../project-task-auth/
172
- # Subagent-db works in: ../project-task-db/
173
- # They cannot interfere with each other's files
174
-
175
- # After all subagents complete, merge in dependency order
176
- MERGE_ORDER=("db" "auth" "api" "tests")
177
- git checkout main
178
- for task in "${MERGE_ORDER[@]}"; do
179
- git merge --no-ff "feature/task-${task}" -m "merge: task-${task}"
180
- git worktree remove "../${REPO_NAME}-task-${task}"
181
- git branch -d "feature/task-${task}"
182
- done
183
- ```
184
-
185
- ### Pattern: Hotfix While Feature Work Continues
186
-
187
- ```bash
188
- # You're in the middle of a long feature
189
- git worktree list
190
- # /Users/you/project [feature/auth-service]
191
-
192
- # Production alert fires — need to hotfix
193
- git worktree add ../hotfix main
194
- cd ../hotfix
195
- # ... fix the bug ...
196
- git commit -m "fix: critical payment timeout in production"
197
- git push origin hotfix/payment-timeout
198
- # PR/merge the hotfix from this worktree
199
-
200
- # Back to feature work
201
- cd ../project # Original feature work untouched
202
- git status # Clean, feature work is exactly where you left it
203
- ```
204
-
205
- ## ClawPowers Enhancement
206
-
207
- When `~/.clawpowers/` runtime is initialized:
208
-
209
- **Worktree Lifecycle Management:**
210
-
211
- ```bash
212
- # Register a worktree
213
- bash runtime/persistence/store.sh set "worktree:task-auth:path" "../project-task-auth"
214
- bash runtime/persistence/store.sh set "worktree:task-auth:branch" "feature/task-auth"
215
- bash runtime/persistence/store.sh set "worktree:task-auth:status" "active"
216
- bash runtime/persistence/store.sh set "worktree:task-auth:created_at" "$(date -u +%Y-%m-%dT%H:%M:%SZ)"
217
-
218
- # List all active worktrees with their status
219
- bash runtime/persistence/store.sh list "worktree:*:status"
220
- ```
221
-
222
- If a session is interrupted, the worktree registry shows which are active and which branches they hold — preventing orphaned worktrees.
223
-
224
- **Conflict Prediction:**
225
-
226
- Before creating parallel worktrees, the framework checks for file overlap:
227
-
228
- ```bash
229
- # For each planned worktree pair, check for overlapping file changes
230
- # High overlap = schedule sequentially; low overlap = safe to parallelize
231
- bash runtime/persistence/store.sh set "worktree:conflict_check:task-auth_vs_task-db" "no_overlap"
232
- ```
233
-
234
- **Cleanup Automation:**
235
-
236
- After merge detection, automatically prompt for worktree cleanup:
237
-
238
- ```bash
239
- bash runtime/feedback/analyze.sh --worktrees
240
- # Output:
241
- # Merged branches with active worktrees:
242
- # - feature/task-auth (merged 3 hours ago) → worktree at ../project-task-auth
243
- # Run: git worktree remove ../project-task-auth && git branch -d feature/task-auth
244
- ```
245
-
246
- ## Anti-Patterns
247
-
248
- | Anti-Pattern | Why It Fails | Correct Approach |
249
- |-------------|-------------|-----------------|
250
- | Checking out same branch in two worktrees | Git prevents this — error on checkout | Each worktree must have a unique branch |
251
- | Never cleaning up worktrees | Disk fills up, confusion about active branches | Cleanup immediately after branch merges |
252
- | `--force` on worktree with uncommitted work | Loses uncommitted changes permanently | Commit or stash before removing |
253
- | Parallel worktrees editing the same file | Merge conflicts on integration | Check file overlap before creating parallel worktrees |
254
- | Forgetting which worktree you're in | Wrong branch gets commits | `git worktree list` before committing |
255
- | Long-lived worktrees diverging from main | Painful rebase/merge on integration | Regularly sync worktrees with `git rebase origin/main` |
256
-
257
- ## Integration with Other Skills
258
-
259
- - Used by `subagent-driven-development` for task isolation
260
- - Used by `dispatching-parallel-agents` for concurrent work
261
- - Used by `finishing-a-development-branch` when cleaning up
@@ -1,281 +0,0 @@
1
- ---
2
- name: validator
3
- description: Multi-round automated validation pipeline for any software project. Runs 14 rounds of checks — compile gates, lint, tests, security scanning, documentation, secrets detection, link verification, spelling, cross-platform compatibility, dependency health, and PR-readiness. Auto-detects project language. Use before publish, deploy, merge, or external PR submission.
4
- version: 1.0.0
5
- requires:
6
- tools: [bash, node, npm]
7
- optional_tools: [trivy, gitleaks, codespell, markdownlint-cli2, eslint, cargo, go, python3]
8
- runtime: false
9
- metrics:
10
- tracks: [rounds_passed, rounds_failed, rounds_warned, total_issues, critical_issues, test_count, test_pass_rate, type_coverage_pct, vulnerability_count]
11
- improves: [code_quality, security_posture, documentation_completeness, publish_readiness]
12
- ---
13
-
14
- # Validator
15
-
16
- ## When to Use
17
-
18
- - Before `npm publish` / `cargo publish` / any package release
19
- - Before merging a PR to your own repo
20
- - Before submitting a PR to an external repo (NVIDIA, Google, CNCF, etc.)
21
- - After a major refactor or dependency update
22
- - On any project — auto-detects language from marker files
23
-
24
- **Skip when:**
25
- - Trivial docs-only changes (run rounds 5, 8 only)
26
- - Quick iteration cycles (run round 0 + 2 only: compile + test)
27
-
28
- ## Quick Start
29
-
30
- ```text
31
- Run the Validator on ~/DevDrive/my-project
32
- ```
33
-
34
- Target specific rounds:
35
-
36
- ```text
37
- Run Validator round 0-2 on my-project (compile + lint + test only)
38
- ```
39
-
40
- PR-readiness for external submission:
41
-
42
- ```text
43
- Run Validator PR-readiness checks on my-project for NVIDIA/NeMo-Agent-Toolkit-Examples
44
- ```
45
-
46
- ## Language Auto-Detection
47
-
48
- Detect project type from marker files. When multiple markers exist, run checks for ALL detected languages.
49
-
50
- | Marker File(s) | Language | Compile | Lint | Test | Security |
51
- |---|---|---|---|---|---|
52
- | `package.json` + `tsconfig.json` | TypeScript | `tsc --noEmit` | ESLint | `npm test` | `npm audit` |
53
- | `package.json` (no tsconfig) | JavaScript | `node --check *.js` | ESLint | `npm test` | `npm audit` |
54
- | `Cargo.toml` | Rust | `cargo check` | Clippy + rustfmt | `cargo test` | `cargo audit` |
55
- | `go.mod` | Go | `go build ./...` | golangci-lint | `go test ./...` | `govulncheck` |
56
- | `pyproject.toml` / `setup.py` | Python | `py_compile` | Ruff + Bandit | pytest | Bandit |
57
- | `Dockerfile` | Docker | `docker build --check` | Hadolint | — | Trivy |
58
- | `foundry.toml` | Solidity | `forge build` | `forge fmt --check` | `forge test` | Slither |
59
- | `*.sh` | Shell | `bash -n` | ShellCheck | — | — |
60
-
61
- ## The 14 Rounds
62
-
63
- Execute in order. Round 0 is a **blocking gate** — if it fails, stop everything.
64
-
65
- ### Round 0 — Compile Gate (BLOCKING)
66
-
67
- If this fails, ALL subsequent rounds are blocked. Fix compile errors first.
68
-
69
- ```bash
70
- # TypeScript
71
- npx tsc --noEmit
72
-
73
- # JavaScript
74
- find . -name "*.js" -not -path "*/node_modules/*" -exec node --check {} \;
75
-
76
- # Rust
77
- cargo check
78
-
79
- # Python
80
- python3 -m py_compile <each .py file>
81
- ```
82
-
83
- **Pass criteria:** Zero compile errors.
84
-
85
- ### Round 1 — Lint
86
-
87
- ```bash
88
- # TypeScript/JavaScript
89
- npx eslint . --ext .ts,.js,.tsx,.jsx 2>&1
90
-
91
- # Rust
92
- cargo clippy -- -D warnings
93
-
94
- # Python
95
- ruff check . 2>&1
96
-
97
- # Go
98
- golangci-lint run ./...
99
- ```
100
-
101
- **Pass criteria:** Zero errors. Warnings are advisory.
102
-
103
- ### Round 2 — Test Suite
104
-
105
- ```bash
106
- # Node.js
107
- npm test
108
-
109
- # Rust
110
- cargo test
111
-
112
- # Python
113
- pytest -v
114
-
115
- # Go
116
- go test ./...
117
- ```
118
-
119
- **Pass criteria:** All tests pass. Report total count and pass rate.
120
-
121
- ### Round 3 — Security Audit
122
-
123
- ```bash
124
- # Node.js
125
- npm audit --audit-level=high
126
-
127
- # Rust
128
- cargo audit
129
-
130
- # Python
131
- pip-audit
132
-
133
- # Container
134
- trivy fs --severity HIGH,CRITICAL .
135
- ```
136
-
137
- **Pass criteria:** Zero HIGH or CRITICAL vulnerabilities. LOW/MODERATE are advisory.
138
-
139
- ### Round 4 — Type Coverage
140
-
141
- ```bash
142
- # TypeScript
143
- npx type-coverage --at-least 90
144
-
145
- # JavaScript (JSDoc)
146
- # Count @param, @returns, @type annotations
147
- grep -r "@param\|@returns\|@type" --include="*.js" -l | wc -l
148
- ```
149
-
150
- **Pass criteria:** ≥90% for TypeScript. For JS, report JSDoc annotation count.
151
-
152
- ### Round 5 — Documentation
153
-
154
- Check that these exist and are non-trivial:
155
- - [ ] README.md (≥50 lines)
156
- - [ ] Version mentioned in README or badge
157
- - [ ] Installation instructions
158
- - [ ] Usage examples with real code
159
- - [ ] License declared (package.json or LICENSE file)
160
- - [ ] CHANGELOG.md (if versioned package)
161
-
162
- **Pass criteria:** All items checked.
163
-
164
- ### Round 6 — Changelog
165
-
166
- - [ ] CHANGELOG.md exists
167
- - [ ] Current version has an entry
168
- - [ ] Entry describes what changed (not just "bug fixes")
169
-
170
- **Pass criteria:** Current version documented.
171
-
172
- ### Round 7 — Secrets Detection
173
-
174
- ```bash
175
- # gitleaks (git history)
176
- gitleaks detect --source . -v 2>&1
177
-
178
- # detect-secrets (current files)
179
- detect-secrets scan . 2>&1
180
- ```
181
-
182
- **Pass criteria:** Zero real secrets. Document false positives (contract addresses, example values) and recommend `.gitleaksignore` entries.
183
-
184
- ### Round 8 — Spelling
185
-
186
- ```bash
187
- codespell --skip="node_modules,dist,.git,package-lock.json,*.min.js" .
188
- ```
189
-
190
- **Pass criteria:** Zero typos in source code and documentation.
191
-
192
- ### Round 9 — Link Verification
193
-
194
- Check all URLs in README.md and documentation:
195
-
196
- ```bash
197
- # Extract URLs and test each
198
- grep -oP 'https?://[^\s\)\"]+' README.md | while read url; do
199
- code=$(curl -o /dev/null -s -w "%{http_code}" "$url")
200
- if [ "$code" != "200" ] && [ "$code" != "301" ]; then
201
- echo "BROKEN: $url → $code"
202
- fi
203
- done
204
- ```
205
-
206
- **Pass criteria:** All links return 200 or 301. Flag example.com/placeholder URLs as advisory.
207
-
208
- ### Round 10 — PR-Readiness (for external submissions)
209
-
210
- - [ ] Conventional commit messages (`feat:`, `fix:`, `docs:`, etc.)
211
- - [ ] DCO sign-off on commits (`git commit -s`)
212
- - [ ] SPDX license headers in source files
213
- - [ ] No merge commits (rebase-clean history)
214
- - [ ] Branch is up-to-date with target
215
-
216
- **Pass criteria:** All items for external PR targets. DCO/SPDX are advisory for own repos.
217
-
218
- ### Round 11 — Cross-Platform Compatibility
219
-
220
- - [ ] No hardcoded absolute paths
221
- - [ ] No macOS-only or Linux-only commands without guards
222
- - [ ] No case-sensitive filename conflicts
223
- - [ ] `engines` field in package.json (Node.js)
224
- - [ ] `.env.example` exists if `.env` is used
225
-
226
- **Pass criteria:** Works on macOS, Linux, and CI runners.
227
-
228
- ### Round 12 — Dependency Health
229
-
230
- ```bash
231
- # All deps pinned (no * or latest)
232
- grep -E '"[\*]"|"latest"' package.json
233
-
234
- # Lock file committed
235
- ls package-lock.json || ls yarn.lock || ls pnpm-lock.yaml
236
-
237
- # Clean install
238
- npm ci --dry-run
239
- ```
240
-
241
- **Pass criteria:** Deps pinned, lock file committed, clean install works.
242
-
243
- ### Round 13 — Summary & Verdict
244
-
245
- Compile results from all rounds:
246
-
247
- ```
248
- ## Validator Report — [Project] v[Version]
249
-
250
- | Round | Check | Result |
251
- |-------|-------|--------|
252
- | 0 | Compile | ✅/❌ |
253
- | 1 | Lint | ✅/⚠️/❌ |
254
- | ... | ... | ... |
255
-
256
- **Verdict:** PASS ✅ / WARN ⚠️ / FAIL ❌
257
- **Score:** X/14 rounds clean
258
-
259
- Blocking issues: [list or "none"]
260
- Advisory warnings: [list or "none"]
261
- ```
262
-
263
- ## Verdicts
264
-
265
- | Verdict | Meaning |
266
- |---------|---------|
267
- | **PASS ✅** | All rounds clean. Safe to publish/merge. |
268
- | **WARN ⚠️** | No blockers but advisory issues exist. Safe to publish, address warnings when convenient. |
269
- | **FAIL ❌** | Blocking issues in Round 0-3. Fix before proceeding. |
270
-
271
- ## Output
272
-
273
- Save the full report to `ops/reports/validator-YYYY-MM-DD-HH-<project>.md` in the workspace.
274
-
275
- ## Tips
276
-
277
- - Run rounds 0-2 frequently during development (fast feedback)
278
- - Run full 14 rounds before any publish or external PR
279
- - Round 7 (secrets) is critical before pushing to public repos
280
- - Round 10 (PR-readiness) only matters for external repo submissions
281
- - Use `--skip-round N` to skip specific rounds when re-running after fixes