@christianirack/brainllm 0.4.0 → 0.5.0

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 (3) hide show
  1. package/README.md +31 -31
  2. package/brain.schema.json +35 -0
  3. package/package.json +8 -8
package/README.md CHANGED
@@ -1,4 +1,4 @@
1
- <p align="center"><img src="https://raw.githubusercontent.com/chrisifm/BRAINLLM/v0.4.0/docs/site/assets/logo.png" width="88" alt="BRAINLLM brain mark"></p>
1
+ <p align="center"><img src="https://raw.githubusercontent.com/chrisifm/BRAINLLM/v0.5.0/docs/site/assets/logo.png" width="88" alt="BRAINLLM brain mark"></p>
2
2
 
3
3
  # BRAINLLM
4
4
 
@@ -13,10 +13,10 @@
13
13
  <img src="https://img.shields.io/badge/built%20with-Rust-dea584?logo=rust&logoColor=white" alt="Built with Rust">
14
14
  <a href="https://chrisifm.github.io/BRAINLLM/"><img src="https://img.shields.io/badge/website-chrisifm.github.io%2FBRAINLLM-0071e3" alt="Website"></a>
15
15
  <a href="https://www.npmjs.com/package/@christianirack/brainllm"><img src="https://img.shields.io/npm/v/@christianirack/brainllm?color=cb3837" alt="npm version"></a>
16
- <a href="https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/LICENSE"><img src="https://img.shields.io/github/license/chrisifm/BRAINLLM" alt="License"></a>
16
+ <a href="https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/LICENSE"><img src="https://img.shields.io/github/license/chrisifm/BRAINLLM" alt="License"></a>
17
17
  </p>
18
18
 
19
- <p align="center"><a href="https://chrisifm.github.io/BRAINLLM/"><img src="https://raw.githubusercontent.com/chrisifm/BRAINLLM/v0.4.0/docs/site/assets/cover.png" alt="BRAINLLM — The web compiler for LLMs. Open the website: chrisifm.github.io/BRAINLLM"></a></p>
19
+ <p align="center"><a href="https://chrisifm.github.io/BRAINLLM/"><img src="https://raw.githubusercontent.com/chrisifm/BRAINLLM/v0.5.0/docs/site/assets/cover.png" alt="BRAINLLM — The web compiler for LLMs. Open the website: chrisifm.github.io/BRAINLLM"></a></p>
20
20
 
21
21
  <p align="center"><b>🌐 <a href="https://chrisifm.github.io/BRAINLLM/">chrisifm.github.io/BRAINLLM</a></b></p>
22
22
 
@@ -31,14 +31,14 @@ code generation, bundling and minification. The agent can operate it through the
31
31
  a confusing diagnostic. The agent can turn that finding into a tested proposal and, when configured, push it
32
32
  automatically to a review branch. Maintainers decide what merges.
33
33
 
34
- **[Visit the site](https://chrisifm.github.io/BRAINLLM/)** · **[Build your first app](#give-your-agent-its-first-task)** · **[Read the spec](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/SPEC.md)** · **[See the benchmarks](#show-me-the-numbers)** · **[Contribute](#build-the-tool-your-agent-needs)**
34
+ **[Visit the site](https://chrisifm.github.io/BRAINLLM/)** · **[Build your first app](#give-your-agent-its-first-task)** · **[Read the spec](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/SPEC.md)** · **[See the benchmarks](#show-me-the-numbers)** · **[Contribute](#build-the-tool-your-agent-needs)**
35
35
 
36
36
  ## A frontend language with the model in mind
37
37
 
38
38
  An agent writing your app needs a contract it can follow and feedback it can act on. BRAINLLM makes that loop
39
39
  explicit: read the spec, generate a component, check it, apply the fix, test the result.
40
40
 
41
- ![An agent reads the BRAINLLM contract, writes a component and uses compiler feedback to validate the app](https://github.com/chrisifm/BRAINLLM/raw/v0.4.0/docs/site/assets/agent-loop.png)
41
+ ![An agent reads the BRAINLLM contract, writes a component and uses compiler feedback to validate the app](https://github.com/chrisifm/BRAINLLM/raw/v0.5.0/docs/site/assets/agent-loop.png)
42
42
 
43
43
  - **Less to guess.** `brain spec` defines the supported language. One component format, a prescribed project
44
44
  structure and generators keep the model working inside a known contract.
@@ -66,7 +66,7 @@ agent has to reconstruct while preserving code that a developer can read and rev
66
66
  | Verification and delivery | Tests, `brain doctor` and the production build make validation part of the workflow |
67
67
  | Tested compiler improvements | Isolated proposals, validation gates and optional review-branch pushes let the agent contribute fixes |
68
68
 
69
- Use [SPEC.md](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/SPEC.md), the served `brain spec` and [EVOLUTION.md](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/docs/EVOLUTION.md) for exact contracts and limits.
69
+ Use [SPEC.md](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/SPEC.md), the served `brain spec` and [EVOLUTION.md](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/docs/EVOLUTION.md) for exact contracts and limits.
70
70
 
71
71
  ## Where it fits in the frontend stack
72
72
 
@@ -125,8 +125,8 @@ stack. Builds run serially on a shared machine. Variance matters: inspect the ra
125
125
  The output strategies differ (including SSR, client rendering and an Astro React island); this comparison measures
126
126
  production build duration for this fixture, not universal app performance or feature equivalence.
127
127
 
128
- Read the [expanded method, versions, payloads and results](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/bench/build/expanded.md). The original three-stack
129
- [historical results](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/bench/build/results/summary.md) remain available unchanged. Use `bench/build/expanded.mjs`
128
+ Read the [expanded method, versions, payloads and results](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/bench/build/expanded.md). The original three-stack
129
+ [historical results](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/bench/build/results/summary.md) remain available unchanged. Use `bench/build/expanded.mjs`
130
130
  with the documented frozen binary and scratch-directory setup to reproduce this seven-stack run.
131
131
 
132
132
  **Ask your agent to challenge the result:**
@@ -149,11 +149,11 @@ numbers support it. Do not upload results or publish anything automatically.
149
149
  Clean up only the temporary resources this run created.
150
150
  ```
151
151
 
152
- A separate [cache-key microbenchmark](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/bench/build/results/compiler-cache-lookup.json) measured **43.3× faster key
152
+ A separate [cache-key microbenchmark](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/bench/build/results/compiler-cache-lookup.json) measured **43.3× faster key
153
153
  lookup** after removing an unnecessary parse: 10,000 lookups of a 1,050-byte input. That number measures this
154
154
  internal operation only. It is not a full-build speedup or a webpack comparison.
155
155
 
156
- The [benchmark report](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/bench/REPORT.md) also tracks agent-generated apps, browser behavior, payload size and
156
+ The [benchmark report](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/bench/REPORT.md) also tracks agent-generated apps, browser behavior, payload size and
157
157
  failure recovery. Historical LLM results describe their frozen model/spec/compiler runs; they do not guarantee
158
158
  that a new prompt or compiler revision will pass. This project is open to developers and researchers who want to
159
159
  investigate how language design and compiler feedback affect agent behavior. Bring a hypothesis and a reproducer.
@@ -177,7 +177,7 @@ what each result actually establishes.
177
177
 
178
178
  Browser tooling and the preview supervisor currently require macOS or Linux. Screenshot baselines need stable browser,
179
179
  font and app-data inputs. Runtime mount-time/heap budgets remain planned; the compiler rejects those settings rather
180
- than pretending to measure them. See the [implemented contract](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/SPEC.md) and [tool internals](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/docs/INTERNALS.md).
180
+ than pretending to measure them. See the [implemented contract](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/SPEC.md) and [tool internals](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/docs/INTERNALS.md).
181
181
 
182
182
  ## Give your agent its first task
183
183
 
@@ -271,11 +271,11 @@ For an existing app, run `brain agent setup` to merge local registrations, or
271
271
  The hook checks the app after `.brain`, `.jsbrain` and `.cssbrain` edits; it does not run tests or download packages.
272
272
  Use `brain mcp --profile minimal` for six core tools, or `--profile backend` for resource/job contracts and generators.
273
273
  With a source installation, change the registration to command `brain` and arguments `["mcp"]`.
274
- Other clients use their own MCP settings. See [the backend/agent contract](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/docs/BACKEND.md).
274
+ Other clients use their own MCP settings. See [the backend/agent contract](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/docs/BACKEND.md).
275
275
 
276
276
  </details>
277
277
 
278
- For an existing frontend, start with the [migration workflow](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/docs/MIGRATION.md). Its read-only planner recognizes
278
+ For an existing frontend, start with the [migration workflow](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/docs/MIGRATION.md). Its read-only planner recognizes
279
279
  React, Angular, Vue, Svelte, Next and Nuxt projects. Port one tested route or component at a time, preserve its
280
280
  behavior and keep a rollback path. It is an incremental workflow, not an automatic transpiler.
281
281
 
@@ -293,7 +293,7 @@ behavior and keep a rollback path. It is an incremental workflow, not an automat
293
293
  ```
294
294
 
295
295
  The compiler handles signal reads and writes. Derived values update when their inputs change. Styles can live in
296
- the same file and are scoped to the component. The [served spec](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/SPEC.md) defines the supported syntax.
296
+ the same file and are scoped to the component. The [served spec](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/SPEC.md) defines the supported syntax.
297
297
 
298
298
  ## Readable intent. Lean browser output.
299
299
 
@@ -301,7 +301,7 @@ An internal function named `back_to_home` can become a short identifier in produ
301
301
  on scope and the other names in the module. Keep meaningful names where the agent reasons about the code; let the
302
302
  compiler shorten eligible bindings in what the browser receives.
303
303
 
304
- This example comes from the [minifier's scope regression tests](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/crates/brain-min/src/mangle.rs):
304
+ This example comes from the [minifier's scope regression tests](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/crates/brain-min/src/mangle.rs):
305
305
 
306
306
  ```js
307
307
  // Before
@@ -336,29 +336,29 @@ BRAINLLM includes protections at several stages of the app loop:
336
336
  tests and diagnostics; failed stages remain visible to the agent.
337
337
 
338
338
  These are implemented mechanisms, not a promise to catch every bug or repair every failure. The agent still needs
339
- to test the user's workflow. See [runtime behavior in the spec](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/SPEC.md#errors-at-runtime),
340
- [component boundaries](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/runtime/brain-guard.js) and [compiler diagnostics](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/crates/brain-check/).
339
+ to test the user's workflow. See [runtime behavior in the spec](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/SPEC.md#errors-at-runtime),
340
+ [component boundaries](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/runtime/brain-guard.js) and [compiler diagnostics](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/crates/brain-check/).
341
341
 
342
342
  ## Build the tool your agent needs
343
343
 
344
344
  **Build with it. Let your LLM make it better.**
345
345
 
346
- ![An LLM turns an observed problem into an isolated fix, runs validation and submits a contribution for human review](https://github.com/chrisifm/BRAINLLM/raw/v0.4.0/docs/site/assets/improvement-loop.png)
346
+ ![An LLM turns an observed problem into an isolated fix, runs validation and submits a contribution for human review](https://github.com/chrisifm/BRAINLLM/raw/v0.5.0/docs/site/assets/improvement-loop.png)
347
347
 
348
348
  A failed compilation can become a regression test. A repeated slow step can become a benchmark. A fix made while
349
349
  building your app can become a contribution that helps the next developer and their agent.
350
350
 
351
351
  Found a compiler gap while building an app? Turn the smallest reproducible case into a regression test and a
352
- contribution. BRAINLLM includes an [evolution workflow](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/docs/EVOLUTION.md) for isolated proposals, validated compiler
352
+ contribution. BRAINLLM includes an [evolution workflow](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/docs/EVOLUTION.md) for isolated proposals, validated compiler
353
353
  binaries and reversible activation. Configured automatic pushes go to review branches; merges require the
354
354
  principal author's review. This is an explicit development workflow, not a compiler that silently rewrites itself
355
- on every build. Read the [agent improvement playbook](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/docs/AGENT-IDEAS.md) for an opt-in prompt and concrete ideas
355
+ on every build. Read the [agent improvement playbook](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/docs/AGENT-IDEAS.md) for an opt-in prompt and concrete ideas
356
356
  to investigate.
357
357
 
358
358
  Human developers and LLM contributors are welcome. Useful first contributions include reproducible compiler bugs,
359
359
  behavior-preserving migration examples, clearer diagnostics and performance improvements with before/after data.
360
360
 
361
- 1. Read [CONTRIBUTING.md](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/CONTRIBUTING.md) and the relevant [internals](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/docs/INTERNALS.md).
361
+ 1. Read [CONTRIBUTING.md](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/CONTRIBUTING.md) and the relevant [internals](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/docs/INTERNALS.md).
362
362
  2. Make a focused change with a regression test and synthetic sample data.
363
363
  3. Run `bash scripts/verify.sh`. It builds the CLI and checks Rust, runtime, tooling, diagnostic fixtures, privacy
364
364
  and repository language. Verification also requires ripgrep.
@@ -369,7 +369,7 @@ Never include app secrets, private source or customer data in a contribution.
369
369
 
370
370
  ## Know what you are adopting
371
371
 
372
- BRAINLLM is **0.4.0** (see [CHANGELOG.md](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/CHANGELOG.md)). Start with a small app or an isolated route and test the behavior you need. Features marked
372
+ BRAINLLM is **0.5.0** (see [CHANGELOG.md](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/CHANGELOG.md)). Start with a small app or an isolated route and test the behavior you need. Features marked
373
373
  as planned in the source specification are not a supported promise; `brain spec` serves the implemented contract.
374
374
  Third-party React, Angular and Vue components do not become BRAINLLM components automatically. Benchmark wins in
375
375
  build time do not imply that every runtime workload is faster.
@@ -378,15 +378,15 @@ build time do not imply that every runtime workload is faster.
378
378
 
379
379
  | You want to… | Start here |
380
380
  |---|---|
381
- | Give an agent the project context | [llms.txt](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/llms.txt) |
382
- | Write an app | [SPEC.md](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/SPEC.md) and `brain spec` |
383
- | Migrate an existing frontend | [Migration guide](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/docs/MIGRATION.md) |
384
- | Improve the compiler while using it | [Evolution workflow](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/docs/EVOLUTION.md) and [agent ideas](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/docs/AGENT-IDEAS.md) |
385
- | Understand the implementation | [Internals](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/docs/INTERNALS.md) and [crates](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/crates/) |
386
- | Inspect examples and reproduce measurements | [Examples](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/examples/) and [benchmarks](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/bench/) |
387
- | Contribute a change | [Contributor guide](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/CONTRIBUTING.md) |
381
+ | Give an agent the project context | [llms.txt](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/llms.txt) |
382
+ | Write an app | [SPEC.md](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/SPEC.md) and `brain spec` |
383
+ | Migrate an existing frontend | [Migration guide](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/docs/MIGRATION.md) |
384
+ | Improve the compiler while using it | [Evolution workflow](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/docs/EVOLUTION.md) and [agent ideas](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/docs/AGENT-IDEAS.md) |
385
+ | Understand the implementation | [Internals](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/docs/INTERNALS.md) and [crates](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/crates/) |
386
+ | Inspect examples and reproduce measurements | [Examples](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/examples/) and [benchmarks](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/bench/) |
387
+ | Contribute a change | [Contributor guide](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/CONTRIBUTING.md) |
388
388
 
389
- Released under the [MIT license](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/LICENSE).
389
+ Released under the [MIT license](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/LICENSE).
390
390
 
391
391
  Created and maintained by **Christian Irack Flores Maldonado**, principal author.
392
392
  [ingenieria.dev](https://ingenieria.dev) · [irack.mx](https://irack.mx) · [christian@irack.mx](mailto:christian@irack.mx)
@@ -398,5 +398,5 @@ Every merged contribution helps the next developer and their agent. Thank you to
398
398
 
399
399
  [<img src="https://avatars.githubusercontent.com/u/1562002?v=4" width="48" height="48" alt="chrisifm">](https://github.com/chrisifm) [chrisifm](https://github.com/chrisifm)
400
400
 
401
- Attribution is refreshed through a reviewed pull request. [How it works](https://github.com/chrisifm/BRAINLLM/blob/v0.4.0/docs/CONTRIBUTORS.md).
401
+ Attribution is refreshed through a reviewed pull request. [How it works](https://github.com/chrisifm/BRAINLLM/blob/v0.5.0/docs/CONTRIBUTORS.md).
402
402
  <!-- contributors:end -->
package/brain.schema.json CHANGED
@@ -30,6 +30,41 @@
30
30
  "default": "/",
31
31
  "description": "Path the app is hosted under: starts and ends with \"/\", e.g. \"/app/\". brain run, brain build and brain start serve every URL under it."
32
32
  },
33
+ "backend": {
34
+ "type": "object",
35
+ "additionalProperties": false,
36
+ "description": "Explicit native backend HTTP origins and opt-in live resource rooms; no credential values.",
37
+ "properties": {
38
+ "origins": {
39
+ "type": "array",
40
+ "minItems": 1,
41
+ "maxItems": 8,
42
+ "uniqueItems": true,
43
+ "items": { "type": "string", "format": "uri", "pattern": "^https?://[^/@?#\\s]+$" },
44
+ "description": "Canonical public HTTP(S) origins without path or trailing slash. Defaults to the origin of PUBLIC_SITE_URL. --origin replaces this list."
45
+ },
46
+ "sockets": {
47
+ "description": "Absent/false disables native live resources. Every socket origin must also occur in the resolved HTTP origins; no implicit CORS expansion.",
48
+ "oneOf": [
49
+ { "const": false },
50
+ {
51
+ "type": "object",
52
+ "additionalProperties": false,
53
+ "required": ["allowedOrigins"],
54
+ "properties": {
55
+ "allowedOrigins": {
56
+ "type": "array",
57
+ "minItems": 1,
58
+ "maxItems": 8,
59
+ "uniqueItems": true,
60
+ "items": { "type": "string", "format": "uri", "pattern": "^https?://[^/@?#\\s]+$" }
61
+ }
62
+ }
63
+ }
64
+ ]
65
+ }
66
+ }
67
+ },
33
68
  "cache": {
34
69
  "type": "object",
35
70
  "properties": {
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@christianirack/brainllm",
3
- "version": "0.4.0",
3
+ "version": "0.5.0",
4
4
  "description": "BRAINLLM: the Rust-powered full-stack web compiler for LLMs and AI agents. Native brain CLI, dev server, optimized builds, PostgreSQL backend and MCP server.",
5
5
  "bin": {
6
6
  "brain": "bin/brain.js"
@@ -15,13 +15,13 @@
15
15
  "node": ">=22"
16
16
  },
17
17
  "optionalDependencies": {
18
- "@christianirack/brainllm-darwin-arm64": "0.4.0",
19
- "@christianirack/brainllm-darwin-x64": "0.4.0",
20
- "@christianirack/brainllm-linux-x64": "0.4.0",
21
- "@christianirack/brainllm-linux-arm64": "0.4.0",
22
- "@christianirack/brainllm-linux-x64-musl": "0.4.0",
23
- "@christianirack/brainllm-win32-x64": "0.4.0",
24
- "@christianirack/brainllm-win32-arm64": "0.4.0"
18
+ "@christianirack/brainllm-darwin-arm64": "0.5.0",
19
+ "@christianirack/brainllm-darwin-x64": "0.5.0",
20
+ "@christianirack/brainllm-linux-x64": "0.5.0",
21
+ "@christianirack/brainllm-linux-arm64": "0.5.0",
22
+ "@christianirack/brainllm-linux-x64-musl": "0.5.0",
23
+ "@christianirack/brainllm-win32-x64": "0.5.0",
24
+ "@christianirack/brainllm-win32-arm64": "0.5.0"
25
25
  },
26
26
  "keywords": [
27
27
  "llm",