Borrowing it
Nothing to install: this file belongs to Sowiedu/Edict. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/Sowiedu/Edict/main/.agent/skills/edict-program-writer/SKILL.mdgit clone --depth 1 https://github.com/Sowiedu/EdictWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/sowiedu/edict/edict-program-writer)<a href="https://agentmods.dev/skills/sowiedu/edict/edict-program-writer"><img src="https://agentmods.dev/badge/skills/sowiedu/edict/edict-program-writer/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/sowiedu/edict/edict-program-writer"><img src="https://agentmods.dev/badge/skills/sowiedu/edict/edict-program-writer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00106 | $0.02409 |
| Opus 5.5 | $0.00042 | $0.00964 |
| Sonnet 5.5 | $0.00021 | $0.00482 |
| Haiku 4.5 | $0.00011 | $0.00241 |
Grade A, and why
edict-program-writer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 199 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Writing Edict Programs
Edict programs are JSON objects — not text files. You produce an AST directly as structured JSON. There is no syntax to learn, only a schema to conform to.
The Loop
- Call
edict_schemato get the JSON Schema (or readschema/edict-schema.json) - Write a program as a JSON AST conforming to the schema
- Call
edict_check(oredict_compile) — if errors come back, fix and resubmit - Call
edict_runto execute the compiled WASM
Program Structure
Every program is a module:
{
"kind": "module",
"id": "mod-myprogram-001",
"name": "myprogram",
"imports": [],
"definitions": [ ... ]
}
The optional schemaVersion field tracks which schema version the AST targets. You don't need to include it — programs without schemaVersion are treated as v1.0 and auto-migrated to the current version by the compiler. If you include it, set it to the current version (check via edict_version).
A module contains definitions: functions (fn), records (record), enums (enum), type aliases (type), and constants (const).
Every module needs a main function that returns Int (exit code).
ID Conventions
Every AST node needs a unique id string. Convention: {kind}-{descriptive-name}-{counter}
| Kind | Prefix | Example |
|---|---|---|
| Module | mod- |
mod-hello-001 |
| Function | fn- |
fn-main-001 |
| Parameter | param- |
param-n-001 |
| Literal | lit- |
lit-zero-001 |
| Call | call- |
call-print-001 |
| Identifier | ident- |
ident-x-001 |
| Binary op | binop- |
binop-add-001 |
| If | if- |
if-check-001 |
| Let | let- |
let-result-001 |
| Match | match- |
match-shape-001 |
| Record def | rec- |
rec-point-001 |
| Enum def | enum- |
enum-shape-001 |
IDs must be globally unique within a module. Duplicate IDs cause a duplicate_id error.
Type System
| Type | JSON representation |
|---|---|
| Basic | { "kind": "basic", "name": "Int" } — also Float, String, Bool |
| Array | { "kind": "array", "element": <TypeExpr> } |
| Option | { "kind": "option", "inner": <TypeExpr> } |
| Result | { "kind": "result", "ok": <TypeExpr>, "err": <TypeExpr> } |
| Named | { "kind": "named", "name": "Point" } — references a record/enum |
| Tuple | { "kind": "tuple", "elements": [<TypeExpr>, ...] } |
| Function | { "kind": "fn_type", "params": [...], "effects": [...], "returnType": <TypeExpr> } |
| Unit | { "kind": "unit_type", "base": "Float", "unit": "usd" } |
| Refined | { "kind": "refined", "id": "...", "base": <TypeExpr>, "variable": "x", "predicate": <Expr> } |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2mo ago First seen · 199 lines · 106 tokens per session scan A 2e81cad690f3
edict-program-writer is a skill published in the GitHub repository Sowiedu/Edict (13 stars, last pushed 4d ago), licensed MIT. It adds 106 tokens to every session and 2,409 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
sounio-pgo
Work on the profile-guided optimization pipeline: counter injection, .sprof file output, strategy promotion, inlining, layout, const-fold/DCE, and register allocation; use when editing any sprint 38–52+ IR optimization files.
epistemic-uncertainty-quantification
Extend Sounio's epistemic types with advanced uncertainty quantification methods, such as confidence intervals, non‑Gaussian distributions, and Dempster‑Shafer evidence combination, positioning the language at the state‑of‑the‑art in measurement science.
sounio-bootstrap
Work on the self-hosted Sounio compiler bootstrap path: self-hosted/main.sio pipeline modes (--lex, --parse, --check, --ir-dump, --native-compile), frontend corpus gates, and bootstrap binary milestones.
sounio-native-codegen
Work on the native x86-64 code generation backend: register allocation, instruction encoding, peephole optimization, frame sizing, ELF output, and the compileirfunction pipeline; use when editing any sprint 52–65+ native/.sio files.
sounio-render
Work on Sounio render examples, website showcase generation, and GPU backend emitters; use when editing examples/render or the website render preview.
sounio-lint
Scan a .sio file for Rust-isms and suggest Sounio-native fixes.