use-ailang
Skill38 starsChanged 8 days agoApache-2.0View on GitHub
Write and run AILANG code with correct syntax. Use when user asks to write AILANG programs, fix AILANG syntax errors, run AILANG code, or needs help with AILANG syntax. Includes version checking, syntax validation, and common patterns.
In this repository85 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- docs/llms.txtllms.txt
- llms.txtllms.txt
- agent-inboxSkill
- ailang-core-triageSkill
- ailang-packagesSkill
- benchmark-managerSkill
- benchmark-runnerSkill
- builtin-developerSkill
- cli-doc-maintainerSkill
- cloud-cascade-debugSkill
- cloud-setupSkill
- codebase-organizerSkill
- collaboration-hubSkill
- coordinator-helperSkill
- design-doc-creatorSkill
- design-spec-auditorSkill
- docs-syncSkill
- eval-analyzerSkill
- eval-gap-finderSkill
- gemini-cli-helperSkill
- github-issue-triageSkill
- headless-runnerSkill
- local-model-onboardingSkill
- local-ollama-evalSkill
- mission-briefSkill
- mission-controlSkill
- mission-loop-changeSkill
- model-managerSkill
- motoko-analyzerSkill
- parser-developerSkill
- post-releaseSkill
- pr-monitorSkill
- prompt-managerSkill
- release-managerSkill
- sonarcloud-triageSkill
- sprint-evaluatorSkill
- sprint-executorSkill
- sprint-plannerSkill
- test-coverage-guardianSkill
- trace-debuggerSkill
- wasm-debuggerSkill
- agent-inboxSkill
- ailang-core-triageSkill
- ailang-packagesSkill
- benchmark-managerSkill
- benchmark-runnerSkill
- builtin-developerSkill
- cli-doc-maintainerSkill
- cloud-cascade-debugSkill
- cloud-setupSkill
- codebase-organizerSkill
- collaboration-hubSkill
- coordinator-helperSkill
- design-doc-creatorSkill
- design-spec-auditorSkill
- docs-syncSkill
- eval-analyzerSkill
- eval-gap-finderSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 0 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 8
Write correct AILANG code using the live MCP server as the canonical reference.
- Very specific: line 124
When the user asks for AILANG code:
- Very specific: line 127
Get current syntax reference: call prompt_get(for_version=...) via MCP, or ailang prompt locally.
- Very specific: line 128
Discover stdlib + examples: before writing: call stdlib_search / examples_for_concept instead of guessing imports.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
AILANG Code Writing
Write correct AILANG code using the live MCP server as the canonical reference.
Quick Start
ailang run --caps IO --entry main solution.ail # Run with effects
ailang repl # Interactive development
ailang check solution.ail # Type-check only
ailang verify solution.ail # Prove contracts with Z3When to Use This Skill
- User asks to write AILANG code, fix syntax errors, or run programs
- User asks about AILANG features, capabilities, or version
- User wants to add Z3 verification or contracts
Source of Truth: the AILANG MCP Server
mcp.ailang.sunholo.com is the canonical, live, version-locked source of truth for AILANG syntax, stdlib, examples, and limitations. Prefer it over any embedded reference in this file.
If the user's harness supports remote MCP (Codex Desktop, Cursor, Cline, Continue, Codex), have them add it once:
{
"mcpServers": {
"ailang-docs": {
"url": "https://mcp.ailang.sunholo.com/mcp/",
"transport": "streamable-http"
}
}
}Or use the one-click button at ailang.sunholo.com.
Then the agent has typed tools to query AILANG knowledge directly:
| Tool | When to call |
prompt_get(for_version, kind="agent") | Need full syntax reference (replaces reading resources/syntax_quick_ref.md) |
stdlib_modules(for_version) | Discover what's available in stdlib |
stdlib_module(name, for_version) | Look up exports of a specific module before importing |
stdlib_search(query, for_version) | Find a function by name/keyword |
examples_for_concept(concept, for_version) | Find a working example of a feature |
example_get(path, for_version) | Read full source of a specific example |
limitations_list(for_version) | Check what AILANG can't do (Y-combinator etc.) before proposing it |
effects_catalog(for_version) | Find which capability flag a function needs |
submit_feedback(...) | File a bug/limitation report mid-session — lands in our triage queue |
Pass for_version=$(ailang prompt --version-active) (or empty for "latest") so the response is guaranteed to match the user's CLI capabilities.
If the harness does NOT support remote MCP — fall back to the local CLI:
ailang prompt # Full teaching prompt (embedded)
ailang prompt --source mcp # Force fresh copy from MCP if reachable
ailang prompt --source embedded # Force embedded copy (eval reproducibility)
ailang docs std/list # Inspect a stdlib module locally
ailang examples search "pattern matching"
ailang mcp status # Check MCP reachability + driftailang prompt --source auto (the default) prefers MCP-served content when reachable AND its served_for matches the CLI version, otherwise silently falls back to embedded. Reproducible eval runs MUST set --source embedded.
Local CLI Reference (always works, even offline)
Core commands
ailang repl # Interactive
ailang run --caps IO,FS --entry main file.ail # Flags BEFORE filename!
ailang check file.ail # Type-check only
ailang verify file.ail # Z3 contract verification
ailang watch file.ail # Auto-reload on change
ailang iface mymodule # Module interface as JSONCritical: ailang run flags MUST come before the filename. ailang run file.ail --caps IO silently ignores the flags.
Examples (already on disk)
ailang examples search "pattern matching"
ailang examples list --tags adt
ailang examples show adt_option --runThree Module Conventions (memorize these — they bite)
1. REPL — no module declaration
ailang repl2. Module file — standard
module myapp/main
import std/io (println)
export func main() -> () ! {IO} = println("Hello, AILANG!")Module path uses underscores, not hyphens (mcp_tools/lib, not mcp-tools/lib).
3. Eval-harness benchmark — must use module benchmark/solution
module benchmark/solution
import std/io (println)
export func main() -> () ! {IO} = println("Benchmark solution")Workflow
When the user asks for AILANG code:
- Check that the agent has MCP access (if applicable):
ailang mcp statusor look formcp.ailang.sunholo.comin the agent's MCP config. - Get current syntax reference: call
prompt_get(for_version=...)via MCP, orailang promptlocally. - Discover stdlib + examples before writing: call
stdlib_search/examples_for_conceptinstead of guessing imports. - Write the code following the prompt's syntax. Prefer pure functions (
! {}) for core logic — these are Z3-verifiable; keep effectful code (! {IO}) as thin wrappers. - Add contracts to pure functions when verification matters (
requires/ensures). Thenailang verify solution.ail. - Type-check before running:
ailang check solution.ail. - Run with the right caps:
ailang run --caps IO,FS --entry main solution.ail.
Available Scripts
scripts/check_version.sh
Print the active AILANG prompt version (uses prompts/versions.json).
.Codex/skills/use-ailang/scripts/check_version.shscripts/validate_code.sh <file.ail>
Type-check an AILANG file via ailang check.
.Codex/skills/use-ailang/scripts/validate_code.sh solution.ailMigration Note (M-AGENT-MCP)
This skill used to embed extensive syntax reference (resources/syntax_quick_ref.md, common_patterns.md, z3_verification_patterns.md). Those files are still on disk for offline use, but prefer the MCP server — it's version-locked to your CLI and updates without skill edits. See docs/guides/agent-mcp.md for the full tool catalog.
Getting Help
- AILANG MCP: https://mcp.ailang.sunholo.com/mcp/ (the canonical reference)
- Documentation: https://ailang.sunholo.com/
- Examples:
ailang examples search ...or examples/ directory - Issues: https://github.com/sunholo-data/ailang/issues — or call
submit_feedbackfrom inside an MCP-connected session - REPL: type
:helpinailang repl
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

