Compression you can prove. Dense mode agents can live in.
Professional token compression for agents: always-on SIGNAL-3, input compression, minified skill payloads, structured terse replies, and checkpointed long sessions.
### At a glance
| Topic | Summary |
| --- | --- |
| **What you get** | Shorter **instructions + replies** in the agent; **checkpoints** (S3) instead of pasting full thread history when you want them. |
| **What you run** | `npx skills add mattbaconz/signal` → **`signal3`** or **`/signal3`**. No clone required for normal install. |
| **What this tree is** | **`skills/`** = source specs you edit. **`gemini-signal/`** · **`claude-signal/`** · **`kiro-signal/`** = mirrored host packages (don’t hand-edit; see [CONTRIBUTING](https://github.com/mattbaconz/signal/blob/HEAD/CONTRIBUTING.md)). |
Protocol entrypoints: [`skills/signal.min.md`](https://github.com/mattbaconz/signal/blob/HEAD/skills/signal.min.md) · symbols [`skills/signal-core.min.md`](https://github.com/mattbaconz/signal/blob/HEAD/skills/signal-core.min.md) · repo [github.com/mattbaconz/signal](https://github.com/mattbaconz/signal) · releases [CHANGELOG.md](https://github.com/mattbaconz/signal/blob/HEAD/CHANGELOG.md)
## 10-second install
```bash
npx skills add mattbaconz/signal
```
Then use Signal in any prompt:
```text
signal3
```
Slash commands are also supported when your host exposes them:
```text
/signal3
```
That is the normal path. **No clone required.**
```mermaid
flowchart TD
A["Want Signal?"] --> B["Fast install: no clone"]
B --> C["npx skills add mattbaconz/signal"]
C --> D["Use signal3 or /signal3 in chat"]
D --> E["Dense replies and protected technical details"]
A --> F["Always-on default"]
F --> G["Clone github.com/mattbaconz/signal"]
G --> H["Run scripts/install-signal-all.ps1 -AlwaysOn"]
H --> I["Normal prompts default to SIGNAL-3"]
A --> J["Proof check"]
J --> K["Run scripts/auto-benchmark.ps1"]
K --> L["Claims come from checked-in JSON"]
```
### Optional always-on setup
Clone only if you want Signal to become a default host rule so users can stop typing `signal3` entirely:
```bash
git clone https://github.com/mattbaconz/signal
cd signal
```
```powershell
powershell -NoProfile -ExecutionPolicy Bypass -File .\scripts\install-signal-all.ps1 -AlwaysOn
```
### Proof in one command
From a clone:
```powershell
powershell -NoProfile -ExecutionPolicy Bypass -File .\scripts\auto-benchmark.ps1
```
Static v0.4 fixtures show **11/11 fidelity-pass rows**, **73.5% median estimated savings**, and **39.8%–89.9% p10–p90 savings** across input-compression and skill-overhead checks. Live output claims are reported separately and must pass fidelity gates before they count.
### Who it is for
Use Signal when you run long coding-agent sessions, keep large `AGENTS.md` / `CLAUDE.md` / `GEMINI.md` files, or need terse replies without turning the assistant into a joke persona. The practical promise is simple: **less repeated context, fewer filler tokens, same protected technical details.**
---
## Demo
**What to trust first:** the v0.4 proof suite separates static input/skill shrinkage from live output-token claims. Static proof currently has **11/11 fidelity-pass rows**, **73.5% median estimated savings**, and **39.8%–89.9% p10–p90 savings** across input-compression and skill-overhead fixtures. Live wins count only when the fidelity gate passes. Details: [Benchmark](#benchmark) · [docs/benchmark-methodology.md](https://github.com/mattbaconz/signal/blob/HEAD/docs/benchmark-methodology.md) · [docs/token-metrics.md](https://github.com/mattbaconz/signal/blob/HEAD/docs/token-metrics.md).
**What’s illustration:** scenarios A–C use `ceil(characters / 4)` (not billed API tokens) — good for shape, not primary “proof” vs hosts that report real tokenizer counts.
**Reproduce:** [benchmark/run.ps1](https://github.com/mattbaconz/signal/blob/HEAD/benchmark/run.ps1) (proof suite, live dry-runs, and long-session runners) · methodology [docs/benchmark-methodology.md](https://github.com/mattbaconz/signal/blob/HEAD/docs/benchmark-methodology.md).
---
## Before / after
| 👤 **Verbose agent** | 🌐 **SIGNAL** |
| ----------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------- |
| “I think the problem might be in `auth.js` around line 47…” | `auth.js:47` · null ref · guard — **~7× fewer tokens** in the scripted benchmark. |
| Paste 10 turns of chat + tool noise into context. | **CKPT atom**: stack, progress, next step — transcript stays out of the window. |
| One giant `SKILL.md` tree + references forever. | **Canonical `.md`** for humans, **`.min.md`** for the agent — **~85%** smaller across the eight main skill pairs ([Benchmark](#benchmark)). |
---
## Install
### Fast install
```bash
npx skills add mattbaconz/signal
```
Use `signal3` or `/signal3` in your agent chat to activate the recommended dense mode for long agentic work. Use `/signal` for the smallest activation surface and `/signal2` for a middle tier.
Global install:
```bash
npx skills add mattbaconz/signal -y -g
```
### Always-on setup
The fast install gives you the skill without cloning. Clone the repo only when you want the helper scripts that write host memory files so normal prompts default to SIGNAL-3.
```bash
git clone https://github.com/mattbaconz/signal
cd signal
```
```powershell
powershell -NoProfile -ExecutionPolicy Bypass -File .\scripts\install-signal-all.ps1 -AlwaysOn -DryRun
powershell -NoProfile -ExecutionPolicy Bypass -File .\scripts\install-signal-all.ps1 -AlwaysOn
```
The dry run prints every file that would change. The real run updates supported host instruction files such as `AGENTS.md`, `CLAUDE.md`, and `GEMINI.md`. Manual templates are documented in [docs/always-on.md](https://github.com/mattbaconz/signal/blob/HEAD/docs/always-on.md).
### Compatibility matrix
| Host | Best path | Auto-activation |
| --- | --- | --- |
| Claude Code | Plugin marketplace or standalone `~/.claude/skills/` copy | Plugin/hooks can remind; slash command still explicit |
| OpenAI Codex | `npx skills add mattbaconz/signal` or copy into `.codex/skills` | `AGENTS.md` can make S3 always-on |
| Gemini CLI | Gemini extension package or repo-root `GEMINI.md` | `GEMINI.md` can make S3 always-on |
| Cursor | Copy skills/rules into Cursor skill/rule paths | `.cursor/rules/signal.mdc` can make S3 always-on |
| Windsurf | Generated `.windsurf/rules/signal.md` | Rules can apply automatically |
| Cline | Generated `.clinerules/signal.md` | Rules can apply automatically |
| GitHub Copilot | Generated `.github/copilot-instructions.md` | Instruction file applies in repo context |
| Kiro | Import from `kiro-signal/skills/` subtree | Per-skill import; activation depends on Kiro workspace |
### Claude Code (app + CLI)
**Plugins only exist in [Claude Code](https://code.claude.com/docs/en/overview)** (the coding agent). The consumer **claude.ai** desktop chat does **not** expose **`/plugin`** — if slash commands like **`/plugin`** are ignored or “unknown,” you’re either in the wrong app or on an old Claude Code build. [Update Claude Code](https://code.claude.com/docs/en/setup) and restart; official troubleshooting: [Discover plugins — “/plugin command not recognized”](https://code.claude.com/docs/en/discover-plugins).
**A) Plugin marketplace (when `/plugin` works)** — in the **Claude Code** session, same input as other slash commands:
```text
/plugin marketplace add mattbaconz/signal
/plugin install signal@signal-suite
```
(`owner/repo` works for GitHub; you can use a full Git URL instead.) Then run **`/reload-plugins`**. Marketplace id is **`signal-suite`** — confirm with **`/plugin marketplace list`**. Details: [`claude-signal/README.md`](https://github.com/mattbaconz/signal/blob/HEAD/claude-signal/README.md).
With the plugin, skills are namespaced: **`/signal:signal`**, **`/signal:signal-commit`**, **`/signal:signal-push`**, … (plugin name **`signal`**, then **`:`**, then folder name under [`claude-signal/skills/`](https://github.com/mattbaconz/signal/tree/HEAD/claude-signal/skills/)).
**B) No `/plugin` — standalone skills (works everywhere Claude loads `~/.claude/skills/`):** clone this repo, run **`scripts/sync-integration-packages.ps1`** once if needed, then:
```powershell
powershell -NoProfile -ExecutionPolicy Bypass -File .\scripts\install-signal-all.ps1
```
That copies [`claude-signal/skills/`](https://github.com/mattbaconz/signal/tree/HEAD/claude-signal/skills/) into **`%USERPROFILE%\.claude\skills\`**. Then use short commands like **`/signal`**, **`/signal-commit`** (see [Commands](#commands)). **Do not** also install the plugin for the same skills.
**C) One-off:** run Claude Code from a terminal with the plugin dir — [`claude-signal/README.md`](https://github.com/mattbaconz/signal/blob/HEAD/claude-signal/README.md) (`claude --plugin-dir ./claude-signal` from the repo).
**Optional:** merge [`templates/claude-CLAUDE.min.md`](https://github.com/mattbaconz/signal/blob/HEAD/templates/claude-CLAUDE.min.md) into your project **`CLAUDE.md`** for thin always-on defaults.
### Always-on SIGNAL-3
Users should not need to type `/signal3` on every prompt. Add the host instruction file once, then prompt normally. The agent treats `signal3`, `SIGNAL-3`, and `/signal3` as the same reset command.
```powershell
powershell -NoProfile -ExecutionPolicy Bypass -File .\scripts\install-signal-all.ps1 -AlwaysOn -DryRun
powershell -NoProfile -ExecutionPolicy Bypass -File .\scripts\install-signal-all.ps1 -AlwaysOn
```
Details and host file map: [`docs/always-on.md`](https://github.com/mattbaconz/signal/blob/HEAD/docs/always-on.md).
---
## Commands
| Command | What it does |
| ---------------- | ---------------------------------- |
| `/signal` | S1 — entry tier |
| `/signal2` | S2 — strong default |
| `/signal3` | S3 — auto-CKPT |
| `/signal-commit` | Stage + conventional commit |
| `/signal-push` | Commit + push |
| `/signal-pr` | Push + PR (`gh`) |
| `/signal-review` | One-line review, severity required |
| `/signal-state` | `.signal_state.md` |
| `/signal-compress` | Compress memory/rules/docs with fidelity gates |
| `/signal-diff` | Summarized changes |
| `/signal-search` | Summarized search |
Tier detail: [Tiers](#tiers).
---
## Repo map
There is **no** legacy top-level `signal/` directory in this repo—ignore older docs that referred to it. This table is what you actually open after cloning.
| Location | What it is | You need it if… |
| ---------------------------------------------------------------------- | ---------------------------------------------------------------------------- | --------------------------------------------------------------------- |
| `[skills/](https://github.com/mattbaconz/signal/tree/HEAD/skills/)` | **Canonical** skill specs: `*.md` (readable) + `*.min.md` (dense) | You install via `npx skills add` or copy skills into an agent |
| `[gemini-signal/](https://github.com/mattbaconz/signal/tree/HEAD/gemini-signal/)`, `[claude-signal/](https://github.com/mattbaconz/signal/tree/HEAD/claude-signal/)` | **Mirrored** host extension layouts (`SKILL.md` per tool) | You ship the Gemini CLI or Claude Code plugin from this tree |
| `[kiro-signal/](https://github.com/mattbaconz/signal/tree/HEAD/kiro-signal/)` | **Kiro** mirror: `skills/` + bundled `references/` with rewritten paths | You import into Kiro IDE — see [docs/kiro.md](https://github.com/mattbaconz/signal/blob/HEAD/docs/kiro.md) |
| `[references/](https://github.com/mattbaconz/signal/tree/HEAD/references/)` | Shared refs (symbols, Karpathy norms, benchmarks, checkpoint notes) | You cite norms or symbols |
| `[templates/](https://github.com/mattbaconz/signal/tree/HEAD/templates/)` | Snippets to merge into a **project’s** GEMINI / CLAUDE files | You integrate SIGNAL into an app repo |
| `[scripts/](https://github.com/mattbaconz/signal/tree/HEAD/scripts/)`, `[benchmark/](https://github.com/mattbaconz/signal/tree/HEAD/benchmark/)` | `shrink.ps1`, `sync-integration-packages.ps1`, `verify.ps1`, proof benchmark runners | You contribute or verify locally ([CONTRIBUTING.md](https://github.com/mattbaconz/signal/blob/HEAD/CONTRIBUTING.md)) |
---
## Why SIGNAL
SIGNAL is not just “be brief.” It is a protocol surface for agents that need dense output, compressed input, durable state, and reproducible evidence. The core bet is that professional compression needs gates and artifacts, not only a terse personality.
```mermaid
flowchart LR
subgraph inputs [Verbose]
V1[Long replies]
V2[Full thread history]
V3[Hedging and filler]
end
subgraph signal [SIGNAL]
S[Symbol grammar]
C[CKPT summaries]
M[Minified skills]
I[Input compression]
end
subgraph outputs [Dense]
D1[Terse pointers]
D2[Checkpoint atom]
D3[~85% smaller payloads]
D4[Reproducible JSON]
end
V1 --> S
V2 --> C
V3 --> S
S --> D1
C --> D2
M --> D3
I --> D4
```
**Cumulative** transcript savings (baseline vs checkpoint-style history) are covered in [benchmark/README.md](https://github.com/mattbaconz/signal/blob/HEAD/benchmark/README.md) (`benchmark/long-session/` after a full clone). **Prompt vs output vs `tokens.total`** — hosts report different scopes; see [docs/token-metrics.md](https://github.com/mattbaconz/signal/blob/HEAD/docs/token-metrics.md).
---
## Signal vs Caveman-style
[Caveman](https://github.com/JuliusBrussee/caveman) is the right comparison point because it made terse agent output easy to install and popular. Its public README currently reports a one-line `npx skills add JuliusBrussee/caveman` install, average output savings around **65%** across Claude API benchmark prompts, and `caveman-compress` input savings around **46%** on memory files. The [Caveman release notes](https://github.com/JuliusBrussee/caveman/releases) also emphasize a reproducible benchmark system and one-command install.
SIGNAL v0.4.0 competes on a different axis: professional dense mode with proof gates.
| Capability | Caveman-style baseline | SIGNAL v0.4.0 |
| --- | --- | --- |
| Output style | Persona/telegraphic compression | Professional dense protocol: terse, no preamble, confidence tokens, templates |
| Input compression | Memory-file compression | `signal-compress` for memory, prompts, rules, and docs with protected-token checks |
| Benchmark controls | Strong terse-control idea | Four arms: `baseline`, `terse-control`, `caveman-style`, `signal` |
| What counts as a win | Reported savings plus validation | No win unless fidelity passes: code, paths, line numbers, errors, versions, and terms preserved |
| Long sessions | Output/input savings | Checkpoint-oriented methodology for cumulative context growth |
| Install posture | One-command install | One-command install plus always-on SIGNAL-3 host rules |
The fair claim is not “SIGNAL always emits fewer output tokens than Caveman.” Caveman is very strong at raw output compression. The stronger SIGNAL claim is: **Signal is easier to trust in professional agent workflows because every benchmark claim must pass a fidelity gate and map back to reproducible JSON.**
---
## Tiers
Use `/signal`, `/signal2`, or `/signal3`.
| Tier | You get | Rough habit savings |
| ------ | ---------------------------------------- | -------------------------- |
| **S1** | Symbols, no preamble, no hedge, terse | ~35% |
| **S2** | S1 + BOOT, aliases, delta-friendly turns | another ~20% on top |
| **S3** | S2 + **auto-checkpoint every 5 turns** | long sessions stay bounded |
---
## Symbol grammar (snippet)
| Symbol | Meaning | Example |
| ------ | --------------------- | -------------------- |
| `→` | causes / produces | `nullref→crash` |
| `∅` | none / remove / empty | `cache=∅` |
| `Δ` | change / diff | `Δ+cache→~5ms` |
| `!` | required / must | `!fix before deploy` |
| `[n]` | confidence 0.0–1.0 | `fix logic [0.95]` |
Full reference: `[skills/signal-core.min.md](https://github.com/mattbaconz/signal/blob/HEAD/skills/signal-core.min.md)`.
---
## Benchmark
v0.4.0 uses a proof-first benchmark methodology. A compression win counts only when fidelity passes: protected code, paths, line numbers, versions, quoted errors, identifiers, and technical terms must remain intact. The generated diagram above is a visual summary; the source of truth is the checked-in JSON and runner.
Benchmark arms:
| Arm | Purpose |
| --- | --- |
| `baseline` | Normal agent/system prompt |
| `terse-control` | Explicit "answer concisely" control |
| `caveman-style` | Telegraphic/persona-style compression control |
| `signal` | Actual SIGNAL skill/defaults |
```mermaid
flowchart LR
P["Scenario fixtures"] --> A["4 arms: baseline, terse-control, caveman-style, signal"]
A --> M["Measure input tokens, output tokens, total tokens, chars"]
M --> F["Fidelity gate"]
F -->|Pass| J["Stable JSON row"]
F -->|Fail| X["No benchmark win counted"]
J --> R["README claim: median plus p10/p90 ranges"]
J --> D["Docs: methodology, metrics, release notes"]
```
**Current static snapshot:** [`benchmark/results/v0.4-static.json`](https://github.com/mattbaconz/signal/blob/HEAD/benchmark/results/v0.4-static.json) reports **11/11 fidelity-pass rows**, **73.5% median estimated savings**, and **39.8%–89.9% p10–p90 savings** across input-compression fixtures and skill-overhead checks. Live output and long-session results are reported separately because provider token accounting varies by host.
Methodology: [`docs/benchmark-methodology.md`](https://github.com/mattbaconz/signal/blob/HEAD/docs/benchmark-methodology.md). Runner: [`benchmark/proof-suite.ps1`](https://github.com/mattbaconz/signal/blob/HEAD/benchmark/proof-suite.ps1).
### What the static proof covers
| Category | Rows | Current result | Source |
| --- | ---: | --- | --- |
| Input compression | 3 | `3/3` fidelity pass; median savings `39.8%` | [`benchmark/fixtures/input-compress.json`](https://github.com/mattbaconz/signal/blob/HEAD/benchmark/fixtures/input-compress.json) |
| Skill overhead | 8 | `8/8` fidelity pass; minified surfaces shrink the loaded skill payload | [`skills/`](https://github.com/mattbaconz/signal/tree/HEAD/skills/) |
| Combined static snapshot | 11 | `11/11` fidelity pass; median savings `73.5%`; p10–p90 `39.8%–89.9%` | [`benchmark/results/v0.4-static.json`](https://github.com/mattbaconz/signal/blob/HEAD/benchmark/results/v0.4-static.json) |
### Output compression
Output benchmarks are intentionally harder than the old “verbose vs terse” demo. They compare four arms:
- `baseline`
- `terse-control`
- `caveman-style`
- `signal`
The default live command runs a one-scenario smoke test and writes ignored local JSON under `benchmark/results/local/`. Full live output runs are opt-in because provider capacity, model version, and token accounting move over time.
```powershell
powershell -NoProfile -ExecutionPolicy Bypass -File .\scripts\auto-benchmark.ps1 -Live
```
Full live output run:
```powershell
powershell -NoProfile -ExecutionPolicy Bypass -File .\scripts\auto-benchmark.ps1 -Live -MaxLiveScenarios 0
```
Do not promote a live output table into the README unless every claimed winning row has `success=true` and a fidelity score.
### Skill pairs (canonical `.md` → `.min.md`) — primary on-disk proof
| Pair | Bytes (≈) | Est. tok (≈) | Shrink |
| ----------------- | ------------------ | ----------------- | -------- |
| signal | 3.3K → 0.9K | ~819 → ~217 | ~74% |
| signal-ckpt | 5.6K → 0.7K | ~1389 → ~163 | ~88% |
| signal-commit | 8.3K → 0.7K | ~2071 → ~178 | ~91% |
| signal-compress | 2.9K → 0.9K | ~717 → ~221 | ~69% |
| signal-pr | 4.7K → 0.5K | ~1177 → ~130 | ~89% |
| signal-push | 3.7K → 0.5K | ~936 → ~131 | ~86% |
| signal-review | 5.5K → 0.6K | ~1378 → ~145 | ~90% |
| signal-state | 2.0K → 0.7K | ~511 → ~163 | ~68% |
| **8 pairs total** | **~36.0K → ~5.3K** | **~8996 → ~1329** | **~85%** |
Min-only helpers (`signal-core`, `signal-diff`, `signal-search`) ≈ **1.6K** bytes (~**389** est. tokens).
### Reading the numbers & multi-turn proof
`[docs/token-metrics.md](https://github.com/mattbaconz/signal/blob/HEAD/docs/token-metrics.md)` — prompt vs output vs `tokens.total`. **Multi-turn / cumulative** protocol benchmark: `[benchmark/long-session/](https://github.com/mattbaconz/signal/tree/HEAD/benchmark/long-session/)` (see `[benchmark/README.md](https://github.com/mattbaconz/signal/blob/HEAD/benchmark/README.md)`) — that is where checkpoint-style savings **compound** vs a growing baseline transcript.
### Reproduce
**Static** (heuristic + skill table printout; no API):
```powershell
powershell -NoProfile -ExecutionPolicy Bypass -File .\scripts\auto-benchmark.ps1
```
**Long-session** (multi-turn; many API calls):
```powershell
powershell -NoProfile -ExecutionPolicy Bypass -File .\benchmark\run.ps1 -Mode LongSession -Quick
```
---
## Architecture
```mermaid
flowchart TB
subgraph protocol [Protocol layer]
P["signal.min.md"]
CORE["signal-core.min.md"]
end
subgraph workflow [Workflow skills]
CKPT["signal-ckpt"]
GIT["commit / push / pr"]
REV["signal-review"]
end
subgraph context [Context tools]
DIFF["signal-diff"]
SRCH["signal-search"]
CMP["signal-compress"]
ST["signal-state"]
end
subgraph persist [Persistence]
STATE[".signal_state.md"]
end
P --> CORE
P --> CKPT
P --> GIT
P --> REV
P --> DIFF
P --> SRCH
P --> CMP
ST --> STATE
```
**Tier ladder:**
```mermaid
flowchart LR
S1["/signal — S1"] --> S2["/signal2 — S2"]
S2 --> S3["/signal3 — S3"]
S1 -.- D1["~35% vs loose chat"]
S2 -.- D2["BOOT · aliases · Δ"]
S3 -.- D3["Auto-CKPT every 5 turns"]
```
---
## Coding norms (Karpathy-style)
**Tiers** shape **assistant chat** (symbols, templates, checkpoints). **Karpathy-style norms** shape **implementation work** (how you edit code and ship commits): orthogonal axes—activating `/signal3` does not replace surgical diffs or explicit assumptions.
Norms in brief (canonical list: `[references/karpathy-coding-norms.md](https://github.com/mattbaconz/signal/blob/HEAD/references/karpathy-coding-norms.md)`):
1. **Assumptions** — explicit over implicit; say when unsure.
2. **Simplicity** — avoid over-engineering.
3. **Surgical diffs** — minimal changes tied to the goal.
4. **Verifiable goals** — reproduce, test, or verify where it matters.
5. **No filler** — skip “here is the code”; show the code.
| Resource | Link |
| -------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Full norms | `[references/karpathy-coding-norms.md](https://github.com/mattbaconz/signal/blob/HEAD/references/karpathy-coding-norms.md)` |
| In skills | `[skills/signal.md](https://github.com/mattbaconz/signal/blob/HEAD/skills/signal.md)`, `[skills/signal-core.min.md](https://github.com/mattbaconz/signal/blob/HEAD/skills/signal-core.min.md)` (`KarpathyNorms`), `[skills/signal-commit.min.md](https://github.com/mattbaconz/signal/blob/HEAD/skills/signal-commit.min.md)` (`followKarpathy`) |
| Host templates | `[templates/gemini-GEMINI.md](https://github.com/mattbaconz/signal/blob/HEAD/templates/gemini-GEMINI.md)`, `[templates/claude-CLAUDE.md](https://github.com/mattbaconz/signal/blob/HEAD/templates/claude-CLAUDE.md)` |
---
## Git workflows & CI
| Skill | Role |
| ------------------------------------------------------------ | ------------------------------------------------------ |
| `[skills/signal-commit.min.md](https://github.com/mattbaconz/signal/blob/HEAD/skills/signal-commit.min.md)` | Stage all, conventional commit (`--draft` / `--split`) |
| `[skills/signal-push.min.md](https://github.com/mattbaconz/signal/blob/HEAD/skills/signal-push.min.md)` | Commit + push |
| `[skills/signal-pr.min.md](https://github.com/mattbaconz/signal/blob/HEAD/skills/signal-pr.min.md)` | Commit + push + `gh pr create` |
**CI:** `[.github/workflows/verify.yml](https://github.com/mattbaconz/signal/blob/HEAD/.github/workflows/verify.yml)` runs `[scripts/verify.ps1](https://github.com/mattbaconz/signal/blob/HEAD/scripts/verify.ps1)` on Windows for `main` and PRs.
---
## Changelog
All releases: [CHANGELOG.md](https://github.com/mattbaconz/signal/blob/HEAD/CHANGELOG.md).
---
## Repository layout (clone root)
```
./
├── skills/ # canonical *.md + *.min.md (edit here; see CONTRIBUTING.md)
├── assets/ # logos, benchmark infographic
├── references/ # symbols, Karpathy norms, benchmarks, checkpoint notes
├── templates/ # Gemini / Claude merge snippets
├── scripts/ # shrink.ps1, verify.ps1, sync-integration-packages.ps1
├── benchmark/ # proof-suite.ps1, fixtures, live benchmark runners
├── gemini-signal/ # Gemini CLI extension (mirrored from skills/)
├── claude-signal/ # Claude Code plugin (mirrored from skills/)
├── kiro-signal/ # Kiro IDE import (mirrored + bundled references/)
├── hooks/
└── GEMINI.md # root context (synced from gemini-signal/)
```
---
## Star History
---
*v0.4.0 — Proof-first compression. Professional dense mode.*
