Field Guide

iteration 1/5

---

---

title: "Libcage: the only pure-C LLM coding agent"

slug: libcage-launch

summary: "Why we built a 21KB zero-dependency C agent for autonomous code repair — and what it's for."

tags: [libcage, coding-agent, self-hosted, release]

date: 2026-08-16

---

Every autonomous coding agent today ships on Python or Node. That's fine on a

developer laptop. It's a dead end everywhere else: the embedded box, the

air-gapped CI runner, the minimal container with no venv you trust.

**Libcage** is the exception. A 21KB binary, zero dependencies — libc and POSIX

sockets only. It calls any OpenAI-compatible `/chat/completions` endpoint, reads

your broken file, applies the fix, compiles, and tests in a loop until it's

green.

```

libcage "Fix the off-by-one in parse()" broken.c \

"cc -o out broken.c" "./out"

iteration 1/5

compiling...

testing...

SUCCESS after 1 iteration(s)

```

Why pure C

  • **No supply chain.** One file, auditable by eye. No npm, no pip, no CVE lottery.
  • **Portable to the edge.** `cc -O2 -std=c11` on Linux, macOS, BSD, minimal containers.
  • **Future-proof.** C outlives every language trend.
  • Three tiers

  • **Libcage** ($29) — the autonomous repair agent.
  • **Libcage Pro** ($99) — CycloneDX SBOM generation + endpoint-policy enforcement.
  • **Libcage Team** ($299, 5-seat) — Pro + a tamper-evident HMAC audit log of every repair.
  • All one-time, local-first. Get it: [Libcage](https://aiautomatedsystems.ca/p/libcage) ·

    [Pro](https://aiautomatedsystems.ca/p/libcage-pro) ·

    [Team](https://aiautomatedsystems.ca/p/libcage-team) ·

    [GitHub](https://github.com/Hardonian/libcage).

    ---

    <!-- hermes-crosslink -->

    **Get the production-ready version**

  • Browse the [AI lab product kits](https://aiautomatedsystems.ca/pricing) (n8n, ComfyUI, local-ops checklists).
  • Want this run *for* you? [Request a done-for-you lab audit](https://aiautomatedsystems.ca/lead).
  • Local-first AI drafting — built for regulated work

    🏠 Run the free Sovereign AI Readiness Score →