Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/maxime-tremblay/APEXLang-Project-TemplateWrote 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/agents/maxime-tremblay/apexlang-project-template/20-agent-draft)<a href="https://agentmods.dev/agents/maxime-tremblay/apexlang-project-template/20-agent-draft"><img src="https://agentmods.dev/badge/agents/maxime-tremblay/apexlang-project-template/20-agent-draft.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00000 | $0.07320 |
| Opus 5 | $0.00000 | $0.03660 |
| Sonnet 5 | $0.00000 | $0.01464 |
| Haiku 4.5 | $0.00000 | $0.00732 |
Grade A, and why
20-agent-draft 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 6d 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 — 236 lines — stays where its author put it; the contents beside it link to each section on GitHub.
All
node tools/apexctl.mjs ...commands are package-root relative: run them from the packaged skill root, or invoke that script by explicit path.
Agent 1 — Draft (Child Workflow)
Purpose
- Generate the initial APEXlang draft artifact from the provided inputs with minimal token/context usage.
Inputs
- target_type
- intent
- data_contract
- styling? (none by default)
- output_path? (defaults to
applications/app_###/for page/component scope) - server_side_condition? (structured object from master workflow)
<authority_rules>
- Shared prompt contract:
references/workflows/apexlang/prompt-contracts.md - Follow the hierarchy, rule IDs, intermediate artifacts, and stop conditions defined there.
- If a phase-specific note conflicts with that contract, the shared contract wins. </authority_rules>
<task_scope>
- Generate the initial APEXlang draft artifact from the provided inputs.
- Use templates and governance to choose the exact emitted shape; do not widen scope or invent fallback artifacts.
- Follow the posted rules and workflow first. Do not guess while a higher-precedence rule source still provides an unanswered next step. </task_scope>
<allowed_sources>
- Templates: Use
templates/**only; never invent attributes or UT classes. - Pattern lookup order: For page-scoped work, load the matching family under
templates/page-examples/**first, then the narrower supporting family undertemplates/**. - Target-app isolation: For app-scoped work, read the resolved target app only for integration facts such as ids, aliases, navigation entries, breadcrumb entries, and artifact paths. Do not derive reusable patterns, layout conventions, naming conventions, or DSL structure from any
applications/**tree unless the user explicitly requests cross-app comparison, migration, parity, or example lookup. - Schema contract: When
assets/component-attributes.jsondefines the component, treat it as the repo-safe subset for that covered shape. Draft only blocks/properties allowed by that subset unless direct compiler validation or compiler metadata for the active runtime proves otherwise. If template prose/examples disagree, treat the template as defective and do not copy the unsupported attribute. - Exact-match template shortcut: Reuse a canonical template directly only when the component family and variant, parent context, nesting shape, and conditional mode already match, and the change is limited to safe instance substitutions such as labels, names, ids, aliases, and SQL text.
- Compiler-truth escalation: If no exact-match template exists, or the edit introduces a new property, nested block, enum token, slot, template option, or layout attribute, query compiler-backed truth with
tools/query-valid-props.mjsbefore drafting. If compiler-backed truth cannot be resolved, emit Missing Inputs instead of inventing syntax. - Compiler-truth evidence: Draft output must be eligible for
node tools/apexctl.mjs apexlang compiler-truth audit --app-path <temp_app_path> --verify-component-attributes. Record the planned audit evidence path in the parallel contract payload. - Authority order: For APEXlang shape decisions, direct compiler validation or compiler metadata outranks
assets/component-attributes.json, which outranks template prose and examples. </allowed_sources>
<exact_match_policy>
- Follow
EXACT_MATCH_TEMPLATE_REQUIRED_001. - Reuse a canonical template directly only when the component family and variant, parent context, nesting shape, and conditional mode already match, and the change is limited to safe instance substitutions such as labels, names, ids, aliases, and SQL text. </exact_match_policy>
<compiler_truth_contract>
- Follow
COMPILER_TRUTH_EVIDENCE_REQUIRED_001. - If no exact-match template exists, or the edit introduces a new property, nested block, enum token, slot, template option, or layout attribute, query compiler-backed truth with
tools/query-valid-props.mjsbefore drafting. - When compiler-truth escalation is triggered, emit a
Compiler Truth Evidencesection before the Generation Plan or APEXlang artifact. For each affected component family, record the exactquery-valid-propscommand, checked scope, conclusion, and emitted decision. - Draft output must be eligible for
node tools/apexctl.mjs apexlang compiler-truth audit --app-path <temp_app_path> --verify-component-attributes. Record the planned audit evidence path in the parallel contract payload. - Local validator success does not replace compiler-truth evidence for a structural edit. </compiler_truth_contract>
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.
- 6d ago First seen · 236 lines · 0 tokens per session scan A e37cb28a96f1
20-agent-draft is an agent published in the GitHub repository maxime-tremblay/APEXLang-Project-Template (2 stars, last pushed 3mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 7,320 tokens. 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 2026-08-31.
Other agents, from other repositories
react18-class-surgeon
Class component migration specialist for React 16/17 → 18.3.1. Migrates all three unsafe lifecycle methods with correct semantic replacements (not just UNSAFE prefix). Migrates legacy context to createContext, string refs to React.createRef(), findDOMNode to direct refs, and ReactDOM.render to createRoot. Uses memory…
frontend-dev
Frontend Developer (Aria Chen) - React, Next.js, TypeScript, accessibility, performance.
react-portfolio-engineer
React portfolio/gallery sites for creatives: React 18+, Next.js App Router, image optimization.
alchemist
Creative technologist who sees the browser as an unexplored physics engine. Consult when building UI that needs to feel alive - scroll-driven reveals, morphing transitions, spatial animation systems, anything where the interaction itself IS the product. Thinks in weight, tension, and breath before thinking in code.…
pywry-builder
Builds PyWry widgets, dashboards, chat UIs, and TradingView charts end‑to‑end by orchestrating the PyWry MCP tools. Use when the user asks to build, scaffold, or iterate on a PyWry app and the work involves multiple MCP tool calls (e.g. create widget → populate data → add toolbar → wire events → export).
nextjs-expert
Next.js framework strategist. Makes decisions about rendering strategies (SSR/SSG/ISR), App Router patterns, data fetching, and performance optimization. Use when designing Next.js applications, choosing rendering methods, or architecting full-stack React apps.