aiSlang Agents Catalog

Generated — do not edit by hand. Rebuild with: python3 tools/catalog/scan_catalog.py && python3 tools/catalog/render_catalog.py.

Scanned examples/ on 2026-07-22T11:59:02+00:00 — 16 agents / multi-agent projects, 26 agents in total.

Each entry is a self-contained aiSlang project: declarative capability-based model selection, prompts, agents, deterministic compute skills, RAG knowledge, MCP tools, evals, and a deploy target — all reconstructed straight from the .ais / aislang.toml source.

Index

orchestrator-agent

multi-module · v0.1.0 · 🤝 multi-agent 📚 RAG 🛠 skills ➡️ pipeline 🔁 reflector loop 🐳 compose ✅ evals

v0.3 Week 20 Mon — Orchestrator agent (multi-module project).

Agents

Skills (deterministic compute — the model never does this work)

Composition

• review_once: code_reviewer
  • in: Coder-emitted Patch + Test coverage markdown; out: ACCEPT or specific revision feedback
• from_locked_design: coder
  • in: 5-section markdown design (Architect's output); out: fenced Rust patch + test coverage bullets
• full_design: use architect.architect → use coder.coder → use code_reviewer.code_reviewer → use tester.tester
  • in: natural-language feature spec for a new resource block; out: Tester's Test plan + cargo test invocation + ACCEPT-or-FAIL verdict
• coder_review_cycle: use coder.coder → use code_reviewer.code_reviewer
  • in: Architect's 5-section design markdown; out: Reviewer's verdict (ACCEPT or revision feedback)
• test_once: tester
  • in: Coder's Patch + Test coverage + Reviewer's verdict; out: Test plan + per-skill recommended invocations
• review_cycle (loop): body coder_review_cycle, until {{ output \| startswith('ACCEPT') \| not }}, max 3

Models

RAG knowledge

• Vector store(s): architect_kb (./knowledge/), reviewer_kb (./knowledge/), coder_kb (./knowledge/), tester_kb (./knowledge/)

Tooling & delivery

• Evals: structured_output (f1 ≥ 80%), verdict_shape (f1 ≥ 80%), rust_patch_format (f1 ≥ 80%), test_plan_shape (f1 ≥ 80%)
• Deploy: compose, compose, compose, compose

Source & docs

• Source: examples/agents/orchestrator/aislang.toml (dir: examples/agents/orchestrator)
• Modules: architect (architect), code_reviewer (code_reviewer), coder (coder), orchestrator, tester (tester)

multi-module-demo

multi-module · v0.1.0 · 🤝 multi-agent 🔌 MCP tools

v0.3 Week 16 Mon scaffold — see the sibling aislang.toml.

Agents

Models

RAG knowledge

• Corpus: README.md, company-info.md

Tooling & delivery

• MCP tools: docs → docker.io/mcp/filesystem:1.0.2, docs → docker.io/mcp/filesystem:1.0.2

Source & docs

• Source: examples/multi-module-demo/aislang.toml (dir: examples/multi-module-demo)
• Modules: architect (architect), coder (coder)

patterns-router-demo

single-file · 🤝 multi-agent 🔌 MCP tools ➡️ pipeline

v0.3 Week 17 Fri — worked router-pattern demo.

Agents

Composition

• handle_billing: billing
  • in: the user's billing question; out: a 1-3 sentence billing reply
• handle_technical: technical
  • in: the user's technical question; out: a 2-4 sentence technical reply
• handle_general: general
  • in: the user's general/greeting question; out: a 1-2 sentence conversational reply

Models

RAG knowledge

• Corpus: README.md, company-info.md

Tooling & delivery

• MCP tools: docs → docker.io/mcp/filesystem:1.0.2
• Prompts: billing_agent.md, general_agent.md, router_classifier.md, technical_agent.md

Source & docs

• Source: examples/patterns/router-demo/router-demo.ais (dir: examples/patterns/router-demo)

patterns-orchestrator-demo

single-file · 🤝 multi-agent 🔌 MCP tools ➡️ pipeline 🐳 compose

v0.3 Week 17 Thu — worked orchestrator-pattern demo.

Agents

Composition

• research_then_summarize: researcher → summarizer
  • in: the user's question — e.g. 'Where is Acme HQ?'; out: one-sentence answer drawn from the knowledge base

Models

RAG knowledge

• Corpus: README.md, company-info.md

Tooling & delivery

• MCP tools: docs → docker.io/mcp/filesystem:1.0.2
• Deploy: compose
• Prompts: researcher.md, summarizer.md

Source & docs

• Source: examples/patterns/orchestrator-demo/orchestrator-demo.ais (dir: examples/patterns/orchestrator-demo)

portfolio-pnl

single-file · 🤝 multi-agent 📚 RAG 🛠 skills ➡️ pipeline 🐳 compose ✅ evals

A local-first, multi-agent portfolio analysis app built on aiSlang. It helps one person understand a personal investment portfolio and make deliberate re-balancing decisions — and it runs entirely on your own machine.

Agents

Skills (deterministic compute — the model never does this work)

Composition

• advise: strategic → orchestrator
  • in: the analyze JSON (analytic + strategic metric blocks); out: a keep/rebalance recommendation with rationale

Models

RAG knowledge

• Vector store(s): strategy_kb (./knowledge/)
• Corpus: 01_antifragility.md, 02_barbell.md, 03_fragility_detection.md, 04_rebalancing.md, 05_reading_the_metrics.md

Tooling & delivery

• Evals: explanation_faithful (f1 ≥ 30%), brief_grounded (f1 ≥ 30%)
• Deploy: compose
• Prompts: orchestrator.md, strategic.md

Source & docs

• Source: examples/portfolio-pnl/portfolio.ais (dir: examples/portfolio-pnl)
• README: examples/portfolio-pnl/README.md
• DESIGN: examples/portfolio-pnl/DESIGN.md

qwen-smoke

single-file · 🤝 multi-agent 🐳 compose

v0.3 Week 18 Fri — worked Qwen-on-Ollama smoke demo.

Agents

Models

Tooling & delivery

• Deploy: compose
• Prompts: coder_sys.md, qwen3_sys.md

Source & docs

• Source: examples/qwen-smoke/qwen-smoke.ais (dir: examples/qwen-smoke)

template-demo

single-file · 🤝 multi-agent 🐳 compose

v0.3 Week 18 Tue — worked dynamic-prompts demo.

Agents

Models

Tooling & delivery

• Deploy: compose
• Prompts: dynamic_sys.md, static_sys.md

Source & docs

• Source: examples/template-demo/template-demo.ais (dir: examples/template-demo)

comment-clamper-agent

single-file · ➡️ pipeline 🐳 compose

v0.4 Week 26 Mon — comment-clamper dogfood agent.

Agents

Composition

• clamp_once: comment_clamper
  • in: one verbose Rust doc-comment / module header; out: the clamped comment, same prefix, code untouched

Models

Tooling & delivery

• Deploy: compose
• Prompts: comment_clamper.md

Source & docs

• Source: examples/agents/comment_clamper/comment_clamper.ais (dir: examples/agents/comment_clamper)

db-agent

single-file · 📚 RAG ➡️ pipeline 🐳 compose ✅ evals

v0.3 Week 20 Wed — DB agent.

Agents

Composition

• brief_once: db
  • in: high-level migration goal (string); out: Feature spec (5-section markdown matching Architect's input shape)

Models

RAG knowledge

• Vector store(s): db_kb (./knowledge/)
• Corpus: 01_current_schemas.md, 02_proposed_sqlite_schema.md, 03_reversibility_contract.md

Tooling & delivery

• Evals: feature_spec_shape (f1 ≥ 80%)
• Deploy: compose
• Prompts: db.md

Source & docs

• Source: examples/agents/db/db.ais (dir: examples/agents/db)

e2e-stack

single-file · ☸️ kubernetes

examples/e2e-stack/main.ais — self-contained K8s stack for the managed-cloud multi-tenant e2e (deploy/eks/40-e2e-two-tenants.sh).

Agents

Models

Tooling & delivery

• Deploy: kubernetes
• Prompts: system.md

Source & docs

• Source: examples/e2e-stack/main.ais (dir: examples/e2e-stack)

free-local

single-file · 🐳 compose

examples/free-local.ais — the "no caps, free local model" demo.

Agents

Models

RAG knowledge

• Corpus: README.md, company-info.md

Tooling & delivery

• Deploy: compose
• Prompts: system.md

Source & docs

• Source: examples/free-local.ais (dir: examples)

rag-demo

single-file · 📚 RAG 🐳 compose

v0.3 Week 19 Wed — worked RAG demo end-to-end against bge-m3.

Agents

Models

RAG knowledge

• Vector store(s): aislang_docs (./knowledge/)
• Corpus: aislang_overview.md, composition_primitives.md, k8s_deployment.md, rag.md

Tooling & delivery

• Deploy: compose
• Prompts: doc_helper.md

Source & docs

• Source: examples/rag-demo/rag-demo.ais (dir: examples/rag-demo)

support-agent (paid)

single-file · 🔌 MCP tools 🐳 compose ✅ evals

examples/support-agent-paid.ais — the v0.1 paid-tier demo variant.

Agents

Models

RAG knowledge

• Corpus: README.md, company-info.md

Tooling & delivery

• MCP tools: docs → docker.io/mcp/filesystem:1.0.2
• Evals: factual_accuracy (f1 ≥ 75%)
• Deploy: compose
• Prompts: system.md

Source & docs

• Source: examples/support-agent-paid.ais (dir: examples)

support-agent

single-file · 🔌 MCP tools 🐳 compose ✅ evals

examples/support-agent.ais — the v0.1 canonical reference example.

Agents

Models

RAG knowledge

• Corpus: README.md, company-info.md

Tooling & delivery

• MCP tools: docs → docker.io/mcp/filesystem:1.0.2
• Evals: factual_accuracy (f1 ≥ 50%)
• Deploy: compose
• Prompts: system.md

Source & docs

• Source: examples/support-agent.ais (dir: examples)

support-agent-k8s

single-file · 🔌 MCP tools ☸️ kubernetes ✅ evals

examples/support-agent-k8s.ais — Phase 2 Week 14 Tue.

Agents

Models

RAG knowledge

• Corpus: README.md, company-info.md

Tooling & delivery

• MCP tools: docs → docker.io/mcp/filesystem:1.0.2
• Evals: factual_accuracy (f1 ≥ 50%)
• Deploy: kubernetes
• Prompts: system.md

Source & docs

• Source: examples/support-agent-k8s.ais (dir: examples)

skill-demo

single-file · 🛠 skills 🔌 MCP tools

v0.3 Week 18 Wed-Thu — worked skill { … } block demo.

Skills (deterministic compute — the model never does this work)

RAG knowledge

• Corpus: notes.md

Tooling & delivery

• MCP tools: docs → docker.io/mcp/filesystem:1.0.2

Source & docs

• Source: examples/skill-demo/skill-demo.ais (dir: examples/skill-demo)

Built by tools/catalog — the source of truth for docs.aislang.org's agents catalog page.