ilo 26.5.0

ilo - the token-minimal programming language AI agents write
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161

name: Rust

on:
  push:
    branches: [ "main", "next" ]
  pull_request:
    branches: [ "main", "next" ]

env:
  CARGO_TERM_COLOR: always
  # Interpreter tests exercise deep recursion (e.g. fib(10)), which can blow
  # the 8 MiB default thread stack on Linux debug builds. 16 MiB is enough.
  RUST_MIN_STACK: 16777216

jobs:
  lint:
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@93cb6efe18208431cddfb8368fd83d5badbf9bfd # v5
    - name: Check formatting
      run: cargo fmt --all -- --check
    - name: Clippy
      run: cargo clippy --workspace --all-targets -- -D warnings
    - name: Set up Python
      uses: actions/setup-python@a26af69be951a213d495a4c3e4e4022e16d87065 # v5
      with:
        python-version: '3.12'
    - name: Install skills-ref (agentskills.io reference validator)
      run: pip install git+https://github.com/agentskills/agentskills.git#subdirectory=skills-ref
    - name: Validate Agent Skills spec conformance
      run: skills-ref validate skills/ilo/
    - name: Enforce modular-skill token budget (cl100k_base approx)
      # Aggressive per-module caps set in ILO-382 (measured baseline + ~50-token
      # headroom). See scripts/check-skill-tokens.ilo for per-module overrides.
      # Aggregate across all modules ≤ 12,500. Uses the built-in `tokcount`
      # bytes/3.4 approximation (within ~5% for natural-language skill files);
      # ILO-47 tracks upgrading to the full tiktoken-rs BPE tokeniser.
      run: cargo run --quiet -- run scripts/check-skill-tokens.ilo
    - name: Shellcheck install.sh
      run: shellcheck -s sh scripts/install/install.sh
    - name: Verify install.sh sha256 verification (offline regression test)
      # Stubs curl and exercises happy / tamper / missing-asset paths so a
      # refactor can't silently drop the supply-chain integrity check.
      run: bash scripts/install/test-install-sh.sh

  skill-validate:
    # Validate skills/ilo/SKILL.md against the agentskills.io reference
    # validator. This catches regressions in cross-host compatibility
    # (Claude Code, Claude Desktop, Codex, etc.) before the Rust tests run.
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@93cb6efe18208431cddfb8368fd83d5badbf9bfd # v5
    - name: Install skills-ref
      run: npm install -g skills-ref
    - name: Validate skill
      run: skills-ref validate skills/ilo/
    - name: Print resolved properties (for debug)
      run: skills-ref read-properties skills/ilo/

  build:
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@93cb6efe18208431cddfb8368fd83d5badbf9bfd # v5
    - name: Install cargo-nextest
      uses: taiki-e/install-action@65851e10cd6c377f11a60e600abc07cb08643468 # v2
      with:
        tool: nextest
    - name: Build
      run: cargo build --verbose
    - name: Verify ai.txt is in sync with SPEC.md
      # build.rs regenerates ai.txt from SPEC.md. If a contributor edited SPEC.md without
      # committing the regenerated ai.txt, the diff is non-empty and CI fails.
      run: git diff --exit-code ai.txt
    - name: Run tests (JUnit XML output)
      run: cargo nextest run --profile ci
    - name: Upload test results to Codecov
      if: ${{ !cancelled() }}
      uses: codecov/codecov-action@e79a6962e0d4c0c17b229090214935d2e33f8354 # v6
      with:
        files: target/nextest/ci/junit.xml
        report_type: test_results
        token: ${{ secrets.CODECOV_TOKEN }}

  cross-platform:
    # Smoke-test that ilo builds on every supported native target.
    # This catches linker / cfg regressions before they reach release builds.
    strategy:
      fail-fast: false
      matrix:
        include:
          - target: x86_64-unknown-linux-musl
            os: ubuntu-latest
            cross: true
          - target: aarch64-unknown-linux-musl
            os: ubuntu-latest
            cross: true
          - target: x86_64-apple-darwin
            os: macos-latest
          - target: aarch64-apple-darwin
            os: macos-latest
          - target: x86_64-pc-windows-msvc
            os: windows-latest

    runs-on: ${{ matrix.os }}

    steps:
      - uses: actions/checkout@93cb6efe18208431cddfb8368fd83d5badbf9bfd # v5

      - name: Install Rust
        uses: dtolnay/rust-toolchain@29eef336d9b2848a0b548edc03f92a220660cdb8 # stable
        with:
          targets: ${{ matrix.target }}

      - name: Install cross
        if: matrix.cross
        run: cargo install cross --git https://github.com/cross-rs/cross

      - name: Build (cross)
        if: matrix.cross
        run: cross build --target ${{ matrix.target }} --no-default-features
        shell: bash

      - name: Build (native)
        if: '!matrix.cross'
        run: cargo build --target ${{ matrix.target }} --no-default-features
        shell: bash

  wasm-smoke:
    # Verify wasm32-wasip1 builds cleanly — no runtime smoke yet (wasmtime
    # install adds ~2 min; tracked as part of ILO-16 follow-ups).
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@93cb6efe18208431cddfb8368fd83d5badbf9bfd # v5

      - name: Install Rust
        uses: dtolnay/rust-toolchain@29eef336d9b2848a0b548edc03f92a220660cdb8 # stable
        with:
          targets: wasm32-wasip1

      - name: Build WASM
        run: cargo build --target wasm32-wasip1 --no-default-features

  coverage:
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@93cb6efe18208431cddfb8368fd83d5badbf9bfd # v5
    - name: Install cargo-llvm-cov
      uses: taiki-e/install-action@65851e10cd6c377f11a60e600abc07cb08643468 # v2
      with:
        tool: cargo-llvm-cov
    - name: Build libilo.a (needed by AOT tests)
      run: cargo build
    - name: Generate coverage
      run: cargo llvm-cov --workspace --lcov --output-path lcov.info
    - name: Upload to Codecov
      uses: codecov/codecov-action@e79a6962e0d4c0c17b229090214935d2e33f8354 # v6
      with:
        files: lcov.info
        fail_ci_if_error: false
      env:
        CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}