codebase-ai 0.3.1 → 0.3.3
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.
- package/LICENSE +21 -0
- package/README.md +44 -4
- package/commands/vibeloop.md +257 -0
- package/dist/index.js +67 -66
- package/dist/index.js.map +1 -1
- package/package.json +1 -1
package/LICENSE
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
1
|
+
MIT License
|
|
2
|
+
|
|
3
|
+
Copyright (c) 2024 ZySec-AI
|
|
4
|
+
|
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
6
|
+
of this software and associated documentation files (the "Software"), to deal
|
|
7
|
+
in the Software without restriction, including without limitation the rights
|
|
8
|
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
|
9
|
+
copies of the Software, and to permit persons to whom the Software is
|
|
10
|
+
furnished to do so, subject to the following conditions:
|
|
11
|
+
|
|
12
|
+
The above copyright notice and this permission notice shall be included in all
|
|
13
|
+
copies or substantial portions of the Software.
|
|
14
|
+
|
|
15
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
16
|
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
17
|
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
|
18
|
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
19
|
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|
20
|
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
|
21
|
+
SOFTWARE.
|
package/README.md
CHANGED
|
@@ -5,6 +5,8 @@
|
|
|
5
5
|
<img src="https://img.shields.io/npm/dm/codebase-ai" alt="npm downloads" />
|
|
6
6
|
<img src="https://img.shields.io/github/license/ZySec-AI/codebase" alt="license" />
|
|
7
7
|
<a href="https://github.com/ZySec-AI/codebase/stargazers"><img src="https://img.shields.io/github/stars/ZySec-AI/codebase?style=social" alt="GitHub stars"></a>
|
|
8
|
+
<a href="https://github.com/ZySec-AI/codebase/actions/workflows/ci.yml"><img src="https://github.com/ZySec-AI/codebase/actions/workflows/ci.yml/badge.svg" alt="CI" /></a>
|
|
9
|
+
<a href="https://securityscorecards.dev/viewer/?uri=github.com/ZySec-AI/codebase"><img src="https://api.securityscorecards.dev/projects/github.com/ZySec-AI/codebase/badge" alt="OpenSSF Scorecard" /></a>
|
|
8
10
|
</p>
|
|
9
11
|
|
|
10
12
|
<p align="center">
|
|
@@ -31,7 +33,7 @@ Seven slash commands give AI a complete workflow: simulate real users in a brows
|
|
|
31
33
|
|
|
32
34
|
---
|
|
33
35
|
|
|
34
|
-
## The
|
|
36
|
+
## The loop
|
|
35
37
|
|
|
36
38
|
Once set up, your entire development loop is:
|
|
37
39
|
|
|
@@ -39,7 +41,13 @@ Once set up, your entire development loop is:
|
|
|
39
41
|
/simulate → /build → /launch
|
|
40
42
|
```
|
|
41
43
|
|
|
42
|
-
|
|
44
|
+
Or if you want zero intervention — one command that runs the entire loop automatically:
|
|
45
|
+
|
|
46
|
+
```
|
|
47
|
+
/vibeloop
|
|
48
|
+
```
|
|
49
|
+
|
|
50
|
+
Here's what each step does:
|
|
43
51
|
|
|
44
52
|
---
|
|
45
53
|
|
|
@@ -155,7 +163,7 @@ The autonomous commands (`/simulate`, `/build`, `/launch`) all read the same man
|
|
|
155
163
|
|
|
156
164
|
---
|
|
157
165
|
|
|
158
|
-
## All
|
|
166
|
+
## All slash commands
|
|
159
167
|
|
|
160
168
|
These live in `.claude/commands/` in your project. Commit this folder to share them with your team.
|
|
161
169
|
|
|
@@ -166,6 +174,22 @@ These live in `.claude/commands/` in your project. Commit this folder to share t
|
|
|
166
174
|
| `/build` | Reads your open GitHub Issues. Picks the most important one. Implements the fix. Tests it. Commits it. Closes the issue. Moves to the next. Repeats. |
|
|
167
175
|
| `/launch` | Checks quality gates (bugs, tests, UX score). If everything passes: bumps version, tags release, merges to main, publishes GitHub Release. |
|
|
168
176
|
| `/review` | Deep code audit. Checks for security vulnerabilities, code quality problems, outdated/vulnerable dependencies, and accessibility issues. Everything goes to GitHub Issues. |
|
|
177
|
+
| `/vibeloop` | **The single command that does everything.** Runs `/simulate → /build → /launch` in a fully autonomous loop until your project is shipped. Zero human intervention required. |
|
|
178
|
+
|
|
179
|
+
### `/vibeloop` — the one command to rule them all
|
|
180
|
+
|
|
181
|
+
If you only remember one command, make it this one:
|
|
182
|
+
|
|
183
|
+
```
|
|
184
|
+
/vibeloop # full autonomous run: simulate → build → launch
|
|
185
|
+
/vibeloop --skip-launch # simulate → build only, stop before release
|
|
186
|
+
/vibeloop --dry-run # full run without committing to main or publishing
|
|
187
|
+
/vibeloop --max-rounds 5 # cap the build loop at 5 rounds (default: 20)
|
|
188
|
+
/vibeloop --sim-count 5 # number of simulated users per cycle (default: 3)
|
|
189
|
+
/vibeloop --version 1.2.0 # pin the release version tag
|
|
190
|
+
```
|
|
191
|
+
|
|
192
|
+
`/vibeloop` runs the full loop repeatedly — simulate real users, fix what breaks, clear the issue backlog, ship the release — without you touching the keyboard. You invoke it once and come back to a shipped, tested, tagged release.
|
|
169
193
|
|
|
170
194
|
---
|
|
171
195
|
|
|
@@ -307,6 +331,22 @@ Zero runtime dependencies. Node.js 20+ only.
|
|
|
307
331
|
|
|
308
332
|
---
|
|
309
333
|
|
|
334
|
+
## Contributing
|
|
335
|
+
|
|
336
|
+
We welcome contributions! Please read [CONTRIBUTING.md](CONTRIBUTING.md) for
|
|
337
|
+
guidelines on how to get started, our commit conventions, and the PR process.
|
|
338
|
+
|
|
339
|
+
Found a security issue? See [SECURITY.md](SECURITY.md) — do not open a public issue.
|
|
340
|
+
|
|
341
|
+
## Changelog
|
|
342
|
+
|
|
343
|
+
See [CHANGELOG.md](CHANGELOG.md) for a full version history.
|
|
344
|
+
|
|
345
|
+
## Code of Conduct
|
|
346
|
+
|
|
347
|
+
This project follows a [Code of Conduct](CODE_OF_CONDUCT.md).
|
|
348
|
+
By participating, you agree to uphold it.
|
|
349
|
+
|
|
310
350
|
## License
|
|
311
351
|
|
|
312
|
-
MIT
|
|
352
|
+
MIT — see [LICENSE](LICENSE) for details.
|
|
@@ -0,0 +1,257 @@
|
|
|
1
|
+
---
|
|
2
|
+
description: Full autonomous loop — simulate → build → launch, repeating until shipped. Zero intervention required.
|
|
3
|
+
argument-hint: [--max-rounds N] [--version X.Y.Z] [--dry-run] [--skip-launch] [--sim-count N]
|
|
4
|
+
model: sonnet
|
|
5
|
+
allowed-tools: Agent, Bash(gh:*), Bash(git add:*), Bash(git commit:*), Bash(git push:*), Bash(git checkout:*), Bash(git pull:*), Bash(git fetch:*), Bash(git stash:*), Bash(git log:*), Bash(git status:*), Bash(git diff:*), Bash(git tag:*), Bash(git rev-parse:*), Bash(git branch:*), Bash(git merge:*), Bash(pnpm:*), Bash(npx:*), Bash(npm:*), Bash(node:*), Bash(uv:*), Bash(curl:*), Read, Write, Edit, Glob, Grep
|
|
6
|
+
---
|
|
7
|
+
|
|
8
|
+
# /vibeloop
|
|
9
|
+
|
|
10
|
+
**The single command that does everything.** Runs simulate → build → launch in a fully autonomous loop until your project is shipped. No human intervention required after invocation.
|
|
11
|
+
|
|
12
|
+
```
|
|
13
|
+
/vibeloop # full run: simulate → build → launch
|
|
14
|
+
/vibeloop --skip-launch # simulate → build only, stop before release
|
|
15
|
+
/vibeloop --dry-run # full run with --dry-run passed to launch (no actual release/merge)
|
|
16
|
+
/vibeloop --max-rounds 5 # cap the build loop at 5 rounds (default: 20)
|
|
17
|
+
/vibeloop --sim-count 5 # number of simulated customers per cycle (default: 3)
|
|
18
|
+
/vibeloop --version 1.2.0 # pin the release version tag
|
|
19
|
+
```
|
|
20
|
+
|
|
21
|
+
---
|
|
22
|
+
|
|
23
|
+
## Arguments
|
|
24
|
+
|
|
25
|
+
```
|
|
26
|
+
$ARGUMENTS
|
|
27
|
+
```
|
|
28
|
+
|
|
29
|
+
Parse from `$ARGUMENTS`:
|
|
30
|
+
- `--max-rounds N` → cap build loop rounds (default: 20)
|
|
31
|
+
- `--version X.Y.Z` → pin release version (passed to /launch)
|
|
32
|
+
- `--dry-run` → no commits to main, no GitHub release (passed to /launch)
|
|
33
|
+
- `--skip-launch` → stop after build loop, do not release
|
|
34
|
+
- `--sim-count N` → customers per simulate cycle (default: 3)
|
|
35
|
+
|
|
36
|
+
---
|
|
37
|
+
|
|
38
|
+
## Phase 0 — Preflight
|
|
39
|
+
|
|
40
|
+
```bash
|
|
41
|
+
gh auth status || { echo "ERROR: gh auth login first."; exit 1; }
|
|
42
|
+
git remote get-url origin || { echo "ERROR: No git remote."; exit 1; }
|
|
43
|
+
gh label list --limit 1 --json name --jq '.[0].name' 2>/dev/null | grep -q "sim" || {
|
|
44
|
+
echo "Labels not found — run /setup first."; exit 1;
|
|
45
|
+
}
|
|
46
|
+
[ -f "docs/PRODUCT.md" ] || { echo "docs/PRODUCT.md missing — run /setup first."; exit 1; }
|
|
47
|
+
```
|
|
48
|
+
|
|
49
|
+
### Load codebase project intelligence
|
|
50
|
+
|
|
51
|
+
```bash
|
|
52
|
+
npx codebase brief 2>/dev/null > /tmp/cb-brief.json || true
|
|
53
|
+
```
|
|
54
|
+
|
|
55
|
+
Read `/tmp/cb-brief.json`. Extract and hold in context:
|
|
56
|
+
- `project.name`, `project.description`
|
|
57
|
+
- `commands.dev`, `commands.test`, `commands.build`
|
|
58
|
+
- `stack.frameworks`, `stack.languages`, `stack.package_manager`
|
|
59
|
+
- `git.default_branch` — verify it's `develop`
|
|
60
|
+
|
|
61
|
+
Ensure on develop and clean:
|
|
62
|
+
```bash
|
|
63
|
+
git fetch origin
|
|
64
|
+
git checkout develop && git pull origin develop
|
|
65
|
+
git status --short # warn if dirty but don't block
|
|
66
|
+
```
|
|
67
|
+
|
|
68
|
+
Print the startup banner:
|
|
69
|
+
```
|
|
70
|
+
╔══════════════════════════════════════════════════════════╗
|
|
71
|
+
║ /vibeloop STARTING ║
|
|
72
|
+
╠══════════════════════════════════════════════════════════╣
|
|
73
|
+
║ Project: [name from brief] ║
|
|
74
|
+
║ Stack: [frameworks from brief] ║
|
|
75
|
+
║ Mode: simulate → build → launch ║
|
|
76
|
+
║ Max rounds: [N] ║
|
|
77
|
+
║ Sim customers:[N] per cycle ║
|
|
78
|
+
║ Launch: [yes | --skip-launch | --dry-run] ║
|
|
79
|
+
╚══════════════════════════════════════════════════════════╝
|
|
80
|
+
```
|
|
81
|
+
|
|
82
|
+
---
|
|
83
|
+
|
|
84
|
+
## Phase 1 — Simulate (seed the issue queue)
|
|
85
|
+
|
|
86
|
+
Run a full simulation cycle to find bugs before building anything.
|
|
87
|
+
|
|
88
|
+
### 1a. Run customer journeys
|
|
89
|
+
|
|
90
|
+
Invoke the full `/simulate` logic as a sub-agent:
|
|
91
|
+
|
|
92
|
+
Use the Agent tool to run one complete simulation cycle (Phase 0 through Phase 6 of /simulate). Pass `--count [sim-count]`. The simulation should:
|
|
93
|
+
- Run `[sim-count]` customer journeys against the live dev server
|
|
94
|
+
- Perform the 9-dimension UX audit
|
|
95
|
+
- Fix all fixable bugs inline with atomic commits
|
|
96
|
+
- Create GitHub issues for all findings
|
|
97
|
+
- Write the cycle summary issue
|
|
98
|
+
|
|
99
|
+
When the simulation cycle completes (one full pass), return to this orchestrator.
|
|
100
|
+
|
|
101
|
+
**Do not run /simulate in its own infinite loop** — vibeloop controls the outer loop. Run exactly one simulate cycle here.
|
|
102
|
+
|
|
103
|
+
### 1b. Count open issues after simulate
|
|
104
|
+
|
|
105
|
+
```bash
|
|
106
|
+
ARCH=$(gh issue list --label "arch" --state open --json number --jq 'length')
|
|
107
|
+
BUGS=$(gh issue list --label "bug" --state open --json number --jq 'length')
|
|
108
|
+
```
|
|
109
|
+
|
|
110
|
+
Print:
|
|
111
|
+
```
|
|
112
|
+
PHASE 1 COMPLETE — Simulate seeded [N] arch issues, [N] bug issues
|
|
113
|
+
```
|
|
114
|
+
|
|
115
|
+
If `$ARCH == 0` and `$BUGS == 0`:
|
|
116
|
+
- Print "No issues found. Project looks clean."
|
|
117
|
+
- Skip Phase 2 and proceed directly to Phase 3.
|
|
118
|
+
|
|
119
|
+
---
|
|
120
|
+
|
|
121
|
+
## Phase 2 — Build loop
|
|
122
|
+
|
|
123
|
+
Run the full `/build` loop to resolve all arch and bug issues found by simulate.
|
|
124
|
+
|
|
125
|
+
### 2a. Outer loop (controlled by vibeloop)
|
|
126
|
+
|
|
127
|
+
This phase repeats until either:
|
|
128
|
+
- All `arch` + `vibekit` labeled issues are closed AND no open `bug` issues remain, OR
|
|
129
|
+
- `--max-rounds` is reached
|
|
130
|
+
|
|
131
|
+
For each round:
|
|
132
|
+
|
|
133
|
+
**Step 1 — Build all arch/vibekit issues:**
|
|
134
|
+
|
|
135
|
+
Use the Agent tool to run one full `/build --once` pass (Phase 0 through Phase 2 of /build, then exit). This implements every open `arch`/`vibekit` issue once without running the inner simulate/poll loop — vibeloop controls that loop here.
|
|
136
|
+
|
|
137
|
+
**Step 2 — Simulate verification:**
|
|
138
|
+
|
|
139
|
+
Use the Agent tool to run one `/simulate --journey-only --count [sim-count]` cycle (customer journeys only, no full UX audit). This verifies the build didn't break anything and may find new bugs.
|
|
140
|
+
|
|
141
|
+
**Step 3 — Check gates:**
|
|
142
|
+
|
|
143
|
+
```bash
|
|
144
|
+
npx codebase scan-only --quiet --sync
|
|
145
|
+
ARCH=$(gh issue list --label "arch" --state open --json number --jq 'length')
|
|
146
|
+
BUGS=$(gh issue list --label "bug,critical" --state open --json number --jq 'length')
|
|
147
|
+
BUGS_HIGH=$(gh issue list --label "bug,high" --state open --json number --jq 'length')
|
|
148
|
+
```
|
|
149
|
+
|
|
150
|
+
Print round summary:
|
|
151
|
+
```
|
|
152
|
+
VIBELOOP ROUND [R] / [max-rounds]
|
|
153
|
+
════════════════════════════════════════════════════
|
|
154
|
+
Arch issues remaining: [N]
|
|
155
|
+
Critical/high bugs: [N]
|
|
156
|
+
Status: [continuing | all clear]
|
|
157
|
+
════════════════════════════════════════════════════
|
|
158
|
+
```
|
|
159
|
+
|
|
160
|
+
If `$ARCH == 0` and `$BUGS == 0` and `$BUGS_HIGH == 0`: break loop → proceed to Phase 3.
|
|
161
|
+
|
|
162
|
+
If round >= `--max-rounds`: print "Max rounds reached. Some issues may remain." → proceed to Phase 3 anyway.
|
|
163
|
+
|
|
164
|
+
Otherwise: increment round, repeat Step 1.
|
|
165
|
+
|
|
166
|
+
### 2b. Final full simulate (pre-launch verification)
|
|
167
|
+
|
|
168
|
+
Before launching, run one final comprehensive simulate cycle:
|
|
169
|
+
|
|
170
|
+
Use the Agent tool to run one complete `/simulate` cycle (all phases). This is the final QA gate — any bugs found here must be fixed before launch can proceed.
|
|
171
|
+
|
|
172
|
+
```bash
|
|
173
|
+
BUGS=$(gh issue list --label "bug,critical" --state open --json number --jq 'length')
|
|
174
|
+
BUGS_HIGH=$(gh issue list --label "bug,high" --state open --json number --jq 'length')
|
|
175
|
+
```
|
|
176
|
+
|
|
177
|
+
If critical or high bugs remain after final simulate:
|
|
178
|
+
- Print "Final simulate found blocking bugs. Running one more build pass."
|
|
179
|
+
- Return to Phase 2a for one more round (hard cap at 3 extra rounds regardless of --max-rounds).
|
|
180
|
+
|
|
181
|
+
---
|
|
182
|
+
|
|
183
|
+
## Phase 3 — Launch
|
|
184
|
+
|
|
185
|
+
Skip this phase entirely if `--skip-launch` was passed. Print "Skipping launch (--skip-launch). Done." and exit.
|
|
186
|
+
|
|
187
|
+
### 3a. Pre-launch gate summary
|
|
188
|
+
|
|
189
|
+
```bash
|
|
190
|
+
CRITICAL=$(gh issue list --label "bug,critical" --state open --json number --jq 'length')
|
|
191
|
+
HIGH=$(gh issue list --label "bug,high" --state open --json number --jq 'length')
|
|
192
|
+
ARCH=$(gh issue list --label "arch" --state open --json number --jq 'length')
|
|
193
|
+
```
|
|
194
|
+
|
|
195
|
+
Print:
|
|
196
|
+
```
|
|
197
|
+
PRE-LAUNCH STATUS
|
|
198
|
+
════════════════════════════════════════════════════
|
|
199
|
+
Critical bugs: [N] [BLOCKED if > 0]
|
|
200
|
+
High bugs: [N] [BLOCKED if > 0]
|
|
201
|
+
Arch issues: [N] [WARNING if > 0]
|
|
202
|
+
════════════════════════════════════════════════════
|
|
203
|
+
```
|
|
204
|
+
|
|
205
|
+
If `$CRITICAL > 0` or `$HIGH > 0`:
|
|
206
|
+
- Print "BLOCKED: Open critical/high bugs prevent launch. Fix them first or run /launch --dry-run to inspect."
|
|
207
|
+
- Exit.
|
|
208
|
+
|
|
209
|
+
### 3b. Execute launch
|
|
210
|
+
|
|
211
|
+
Use the Agent tool to run the full `/launch` logic (all phases of /launch). Pass:
|
|
212
|
+
- `--version [version]` if `--version` was specified
|
|
213
|
+
- `--dry-run` if `--dry-run` was specified
|
|
214
|
+
|
|
215
|
+
The launch sub-agent will:
|
|
216
|
+
- Run all gate checks (bugs, tests, UX scores, branch cleanliness)
|
|
217
|
+
- Generate `docs/RELEASE-NOTES.md`
|
|
218
|
+
- Create the GitHub release and tag
|
|
219
|
+
- Merge `develop` → `main`
|
|
220
|
+
- Rotate the milestone
|
|
221
|
+
- Refresh the codebase manifest
|
|
222
|
+
|
|
223
|
+
---
|
|
224
|
+
|
|
225
|
+
## Phase 4 — Final Summary
|
|
226
|
+
|
|
227
|
+
```
|
|
228
|
+
╔══════════════════════════════════════════════════════════╗
|
|
229
|
+
║ /vibeloop COMPLETE ║
|
|
230
|
+
╠══════════════════════════════════════════════════════════╣
|
|
231
|
+
║ Simulate cycles: [N] ║
|
|
232
|
+
║ Build rounds: [N] ║
|
|
233
|
+
║ Issues implemented: [N] ║
|
|
234
|
+
║ Bugs fixed inline: [N] ║
|
|
235
|
+
║ Version released: [vX.Y.Z | --dry-run | skipped] ║
|
|
236
|
+
║ develop → main: [merged | --dry-run | skipped] ║
|
|
237
|
+
╚══════════════════════════════════════════════════════════╝
|
|
238
|
+
|
|
239
|
+
Next steps:
|
|
240
|
+
• Check GitHub releases for the release notes
|
|
241
|
+
• Run /vibeloop again next sprint to start a new cycle
|
|
242
|
+
• Use /simulate for targeted UX testing
|
|
243
|
+
• Use /build --issue N to fix a specific issue
|
|
244
|
+
```
|
|
245
|
+
|
|
246
|
+
---
|
|
247
|
+
|
|
248
|
+
## Ground Rules
|
|
249
|
+
|
|
250
|
+
1. **One agent per phase** — simulate, build, and launch each run as isolated sub-agents via the Agent tool
|
|
251
|
+
2. **vibeloop controls the outer loop** — do not let /build or /simulate run their own infinite loops; vibeloop orchestrates timing
|
|
252
|
+
3. **Never force push** — if git state is broken, investigate before acting
|
|
253
|
+
4. **Hard stop on launch blockers** — critical/high bugs always block Phase 3, no overrides
|
|
254
|
+
5. **Atomic commits throughout** — every fix in every sub-agent must be `git add [specific files]`, never `git add .`
|
|
255
|
+
6. **Dry-run is always safe** — `--dry-run` must propagate to all sub-agents and never touch `main` or create releases
|
|
256
|
+
7. **Max-rounds is a safety net** — if hit, proceed to launch anyway (with warnings in release notes about remaining issues)
|
|
257
|
+
8. **Always on develop** — vibeloop never switches away from develop except for the final merge to main in /launch
|