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

-- check-skill-tokens.ilo
--
-- Enforce the modular-skill token budget.
--
-- Each skills/ilo/ilo-*.md module must encode to <= 1,000 tokens under
-- cl100k_base. Modules currently above the baseline carry an explicit
-- per-module override baked into `limit-for` below (measured baseline +
-- ~50-token headroom, ILO-382). Growth past an override requires an
-- editorial trim or a module split — not a bump.
-- The aggregate across all modules must be <= 12,500.
--
-- `tokcount` uses a bytes/3.4 approximation for cl100k_base (correct
-- within ~5% for natural-language skill files). A follow-up (ILO-47)
-- will replace the stub with a real BPE tokeniser.
--
-- Run locally with: ilo run scripts/check-skill-tokens.ilo

-- Per-module override caps (bytes/3.4 stub measurement + ~60-token headroom).
-- Caps are set for the bytes/3.4 approximation; they will be tightened once
-- the real tiktoken-rs BPE tokeniser lands (ILO-47 follow-up). Modules not
-- listed default to 1000. Stub measurements (2026-05-22):
--   ilo-language:1410  ilo-builtins-core:995  ilo-builtins-math:1323
--   ilo-builtins-io:1834  ilo-builtins-text:1114  ilo-agent:1330
limit-for name:t>n
  ?name{"ilo-language":1550
        "ilo-builtins-core":1060
        "ilo-builtins-math":1390
        "ilo-builtins-io":2550
        "ilo-builtins-text":1300
        "ilo-agent":1600
        _:1000}

-- Ordered list of skill module names (matches Python predecessor exactly).
skill-names>L t
  ["ilo-language",
   "ilo-language-records",
   "ilo-builtins-core",
   "ilo-builtins-math",
   "ilo-builtins-io",
   "ilo-builtins-text",
   "ilo-errors",
   "ilo-tools",
   "ilo-engines",
   "ilo-agent",
   "ilo-examples",
   "ilo-edit-loop"]

-- process-content name:t content:t > n
-- Count tokens in content, print result row, return token count.
process-content name:t content:t>n
  tokens = flr (tokcount content)
  lim = limit-for name
  over = >tokens lim
  flag = ?h over +(+" OVER (cap " str lim) ")" ""
  col1 = padr name 22
  col2 = padl (str tokens) 5
  prnt +(+"  " col1) +col2 +" tokens" flag
  tokens

-- check-module dir:t name:t > R n t
-- Reads a skill file, delegates token counting to process-content.
-- Returns Ok(tokens) on success, Err(message) when the file is missing.
check-module dir:t name:t>R n t
  path = +(+(+dir "/") name) ".md"
  rc = rd path
  ?rc{~content:~process-content name content
      ^er:^+"ERROR: missing skill file: " path}

-- pair-failed pair:L _ > b
-- True when a (name, R n t) pair is missing or over its token cap.
pair-failed pair:L _>b
  name = at pair 0
  r = at pair 1
  ?r{^_:true;~v:>v (limit-for name)}

-- main: aggregate check. Exits 1 (via Err return) if any module is
-- over-budget or the total exceeds 12,500.
main>_
  names = skill-names()
  dir = "./skills/ilo"
  results = map (n:t>R n t; check-module dir n) names
  total = flr (fld (acc:n r:R n t>n; ?r{~v:+acc v;^_:acc}) results 0)
  col1 = padr "TOTAL" 22
  col2 = padl (str total) 5
  prnt +(+"  " col1) +col2 " tokens"
  over-total = >total 12500
  over-total{prnt +(+"ERROR: total " str total) " exceeds aggregate budget 12500"}
  pairs = zip names results
  failures = flt pair-failed pairs
  bad = |(>len failures 0) over-total
  bad{ret ^"budget exceeded"}