marfrit 3e57824684 router: classify_model heuristic + 31-case corpus (Phase 5 commit #1)
Phase 5 commit #1 per docs/PHASE5.md §11. Pure-Lua per-request model
routing — no IO, no LLM probe in v1.

router.classify_model(text, cfg) -> (model_name | nil, class_label):
  1. classify_class(text) walks heuristics in priority order:
       code class:
         - triple-backtick fence anywhere
         - "traceback" / "stacktrace" / "stack trace" (ci)
         - "error:" / "exception:" in first 60 chars (ci)
         - path-with-code-extension token (.py/.lua/.c/.js/.go/.rs/.cpp/.h/.ts)
         - 5+ lines with indented content (looks like a paste)
       reasoning class (requires text >= 15 chars to skip bare keywords):
         - "explain" / "why " / "how does" / "compare" (ci)
         - "?" + length > 100 chars
       default class: everything else
  2. Map class via cfg.routing.classes[class] → model name (or nil = keep current).
  3. Return (model_name_or_nil, class_label).

ALWAYS evaluates regardless of cfg.routing.auto — caller (repl.ask_ai
in commit #3) gates on the flag. This separation lets `:route check`
introspect the heuristic even when routing is off (N1).

M._classify_class exposed for testing.

Test corpus (test_router_model.lua, 31 cases):
  - 13 code-class positives (fence, traceback, paths, multi-line paste)
  - 6 reasoning-class positives (explain/why/how does/compare/?+length)
  - 8 default-class (short queries, bare keywords below 15-char threshold,
    non-code paths like .md/.txt)
  - 3 model-mapping cases (code→"deep", reasoning→"cloud", default→nil)
  - 1 R-N2 default test: classes.reasoning=nil → reasoning text yields
    nil model override (heuristic still fires, no swap)
  - All 31 pass; 15-char threshold catches "how does ASLR work?" without
    false-positive on bare "explain".

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-13 11:17:22 +00:00

aish

aish — AI-augmented conversational shell.

A single REPL that interleaves shell command execution and language-model conversation, backed by a llama.cpp HTTP broker. Implementation is LuaJIT 2.x with FFI bindings to libcurl, GNU readline, and libc — no C extensions, no build step, one source tree.

Why

Three flows that currently live in three windows fold into one:

  1. "Run this command and show me the output" — fast feedback loop, no copy-paste between terminal and chat.
  2. "Explain or write code based on the output we just looked at" — exec output is automatically injected into the model's context.
  3. "Plan and execute a multi-step task with confirmation gates" — landing in Phase 3 as Chuck Norris autonomous mode.

aish is not a wrapper around bash. It's a first-class interactive environment where the shell is one of several execution channels.

Status

Component State
Repository skeleton in this commit
Phase 0 manifest docs/PHASE0.md — locked
Phase 0 implementation 🔜 next session
Phase 1+ 📋 enumerated in PHASE0.md §11

Every module file currently raises not implemented (Phase 0 pending) when called. luajit main.lua fails loudly at the first un-implemented function, never silently.

Quick orientation

Read this If you want to know
docs/PHASE0.md §12 What aish is and what Phase 0 ships
docs/PHASE0.md §3 Technology decisions (LuaJIT, FFI, readline, libcurl, llama.cpp)
docs/PHASE0.md §4 Directory layout — these file names are stable across all phases
docs/PHASE0.md §5 How input is dispatched (meta / shell / AI)
docs/PHASE0.md §6 Broker contract: /v1/chat/completions, CMD: extraction
docs/PHASE0.md §10 Config schema and resolution order
docs/PHASE0.md §11 Phase sequence (what lands when)
docs/PHASE0.md §13 Open questions, tracked per phase
CLAUDE.md Project conventions for AI-assisted contributors

Directory layout

aish/
├── main.lua              # entry point
├── repl.lua              # readline loop, dispatch, prompt
├── broker.lua            # llama.cpp HTTP client
├── router.lua            # input classifier (meta/shell/AI)
├── executor.lua          # command exec + CMD: extraction
├── context.lua           # in-memory turn history
├── history.lua           # disk persistence (Phase 1+)
├── safety.lua            # destructive-op gate (Phase 3+)
├── renderer.lua          # output formatting
├── config.lua            # default model registry + preferences
├── ffi/
│   ├── curl.lua          # libcurl easy interface
│   ├── readline.lua      # GNU readline
│   ├── pty.lua           # forkpty (Phase 1+)
│   └── libc.lua          # chdir, errno, strerror
└── docs/
    └── PHASE0.md         # locked substrate

Build / runtime dependencies

System packages (Debian / ALARM / Arch names):

  • luajit (>= 2.0)
  • libcurl4 / libcurl-openssl-3 runtime
  • libreadline8 runtime
  • libc6 runtime (always present)

No compilation, no luarocks, no make. Just luajit main.lua.

Running

Once Phase 0 ships:

luajit main.lua                          # uses ~/.config/aish/config.lua
luajit main.lua --config ./config.lua    # explicit config path
AISH_CONFIG=/path/to/config.lua luajit main.lua

Config resolution order is documented in docs/PHASE0.md §10.

Configuration

config.lua is a Lua file returning a single table. The committed config.lua in this repo is both the canonical example and the development-fallback config (lowest precedence). Copy it to ~/.config/aish/config.lua and edit endpoints to your local llama.cpp servers, or point AISH_CONFIG at your own.

The default endpoints assume mfritsche's home network:

  • fastdirac.fritz.box:8081 (Qwen2.5-Coder-7B q4 8k ctx)
  • deepdirac.fritz.box:8080 (Qwen2.5-Coder-7B q4 32k ctx)
  • cloudhossenfelder.fritz.box:8082 (forwards to OpenRouter)

Replace these with your own llama.cpp endpoints if you're not on that LAN.

License

Not yet selected. Default-private until decided.

Project conventions

See CLAUDE.md for contribution conventions, commit style, and the phase-loop discipline this project follows.

S
Description
AI-augmented conversational shell — LuaJIT REPL with llama.cpp broker, shell executor, and routed AI inference.
Readme MIT 2.2 MiB
Languages
Lua 99.8%
Shell 0.2%