Skip to content

hashgraph-online/awesome-codex-plugins / plugins/alpoxdev/hypercore/skills/hyper-decision

hyper-decision

Unverified

Use this skill when the user wants a decision API in software: a model call that returns a typed judgment (a yes/no probability, one option from a closed set, an ordered score, or an image box) instead of text, through TypeSafe Jev, OpenAI Decisions, Perplexity, Liquid, Cloudflare Clef, the Vercel AI SDK or AI Gateway, OpenRouter, or an emulation on a general LLM. It decides fit, writes typed questions, calling code, request templates, eval plans, audits of label-only LLM calls, and route setup. Do not use for chat or code generation, model training, moderation policy writing, or general prompt writing.

FileKindSize
SKILL.ko.mdmarkdown27.6 KB
SKILL.mdskill md25.0 KB
assets/evals/hyper-decision-cases.jsonlother36.1 KB
assets/templates/call-aisdk.tsscript9.2 KB
assets/templates/call-direct.tsscript14.5 KB
assets/templates/call-openai.tsscript14.6 KB
assets/templates/call-sdk.pyscript10.1 KB
assets/templates/call-sdk.tsscript9.2 KB
assets/templates/constants.template.tsscript4.2 KB
assets/templates/eval-cases.template.jsonlother3.2 KB
assets/templates/request-openai.jsonconfig2.5 KB
assets/templates/request.jsonconfig2.3 KB
assets/templates/routing-table.template.jsonconfig1.5 KB
references/ecosystem.ko.mdmarkdown7.4 KB
references/ecosystem.mdmarkdown6.6 KB
references/eval-and-audit.ko.mdmarkdown35.0 KB
references/eval-and-audit.mdmarkdown32.5 KB
references/mode-procedures.ko.mdmarkdown26.2 KB
references/mode-procedures.mdmarkdown24.9 KB
references/official/decision-providers.ko.mdmarkdown43.7 KB
references/official/decision-providers.mdmarkdown42.6 KB
references/official/jev-platform.ko.mdmarkdown34.8 KB
references/official/jev-platform.mdmarkdown32.3 KB
references/official/openai-decisions.ko.mdmarkdown20.7 KB
references/official/openai-decisions.mdmarkdown19.9 KB
references/patterns.ko.mdmarkdown20.2 KB
references/patterns.mdmarkdown18.9 KB
references/providers.ko.mdmarkdown30.3 KB
references/providers.mdmarkdown28.9 KB
references/question-design.ko.mdmarkdown34.9 KB
references/question-design.mdmarkdown31.8 KB
rules/modes-and-routing.ko.mdmarkdown12.4 KB
rules/modes-and-routing.mdmarkdown11.4 KB
rules/safety-and-boundaries.ko.mdmarkdown12.1 KB
rules/safety-and-boundaries.mdmarkdown10.7 KB
scripts/check-decision-request.shexecutablescript40.6 KB

references/ecosystem.md

View on GitHub
# Ecosystem

Where this package sits, as of 2026-11-08: the vendor-published agent skill for TypeSafe Jev, the community skills around decision models, and what this package delivers. Every community item is labeled **community, not required, inspiration only**. Nothing here is a dependency.

## Contents

- Refresh Policy
- The vendor-published skill
- Community skills
- What this package delivers
- How to read this landscape
- Unverified items
- Sources

## Refresh Policy

- last_verified_at: 2026-11-08
- refresh_when:
  - the vendor skill ships a new release or changes its install routes
  - a community skill listed here is archived, renamed, or moved
  - another vendor publishes an official decision-API skill
- supports_rules:
  - rules/modes-and-routing.md

## The vendor-published skill

TypeSafe publishes one skill, `typesafe-ai`, in the `typesafe-ai/skills` repository. On 2026-11-08 the repository held a single `SKILL.md` (288 lines) and a license, with no references or scripts; the last release was v0.6.2 on 2026-10-23, a near-complete rewrite of v0.5.7 (2026-09-12).

Install routes on the vendor page; the page asks for exactly one, because a stale copy is what makes an agent invent request or response fields:

```bash
claude plugin marketplace add typesafe-ai/skills
claude plugin install typesafe@typesafe-ai
npx skills add typesafe-ai/skills --skill typesafe-ai
```

What v0.6.2 teaches, and where this package adopted it:

| Vendor guidance | Adopted in |
|---|---|
| Judgments from the model, decisions in code: keep the decision as a truth table and ask only for the judgments | `../rules/modes-and-routing.md` §2 |
| The human test: a capable person with only this state and question could answer at once | `../rules/modes-and-routing.md` §2, `./question-design.md` |
| Confidence is rarely the gate; act on the selected option's probability | `../rules/safety-and-boundaries.md` §5, `./patterns.md` |
| Yellow flags ("prompt gymnastics"): criteria naming test items, a clause per failing case, thresholds fitted to cases, tests asserting exact probabilities, code silently overriding answers | `./eval-and-audit.md` |
| One constants module holding questions, options, criteria, and thresholds as the review surface | `../assets/templates/constants.template.ts` |
| Open the live docs while working, and say so when a page cannot be read | `../rules/safety-and-boundaries.md` §10 |

## Community skills

Each row was opened on 2026-11-08 by reading its `SKILL.md`; the last-commit date is from the repository.

| Family | Name and repository | Last commit | What it does |
|---|---|---|---|
| Design guidance | `system-one`, magnus919/agent-skills | 2026-10-15 | Design, integrate, evaluate, and self-host System One decision models across several vendors |
| Design guidance | `building-with-decision-models`, aaddrick/building-with-decision-models | 2026-10-10 | Vendor-neutral successor of `building-with-typesafe-jev`, one provider file per vendor |
| Design guidance | `building-with-typesafe-jev`, aaddrick/building-with-typesafe-jev | 2026-09-27 | Jev intent-to-shape map, API reference, patterns, six eval cases |
| Vendor gateway | `openrouter-decisions`, OpenRouterTeam/skills | 2026-10-29 | Implements the OpenRouter Decisions API: primitives, state, model pinning, threshold probes |
| Framework | `langgraph-decision-models`, langchain-ai/langchain-skills | 2026-09-25 | Replaces LangGraph routing with decision-model classifiers |
| Question generation | `grill-jev`, grgy078033/grill-jev; Blue-And-White/Grill-Jev; notque/vexjoy-agent | 2026-09-20 to 2026-10-30 | Ask a decision model at each decision point of a plan or task |
| Classification | `jev-triage`, wuyoscar/jev-skill (and several forks) | 2026-09-22 | Inbox and ticket triage with a smoke-test pilot |
| Evaluation | `jev-eval`, OneWave-AI/claude-skills; BarakChamo/jev-kit; wuyoscar/jev-skill | 2026-09-22 to 2026-09-29 | Labeled sets, criteria and threshold sweeps, confident-wrong hunting |
| Evaluation | `jev-avaliar`, ia-automation/How-to-use-JEV | 2026-10-01 | Portuguese-language domain evaluation with a frozen held-out set |
| Audit | `jev-audit`, OneWave-AI/claude-skills; idogoldd/skills; magicbeansai/jev-audit; chaseai-yt/jev-audit | 2026-09-19 to 2026-10-07 | Find label-only LLM calls or heuristic code and rank replacements |

## What this package delivers

- Files, not advice: request JSON per dialect, a constants file, caller code per route and language, a template set, an eval case skeleton, and a read-only audit table.
- One contract across vendors: neutral `boolean`, `choice`, `score`, and `box` types, a dialect map, and four route tiers that say whether a probability comes back.
- An offline checker per dialect that needs no key and no network.
- Volatile facts in dated references only, refreshed after 30 days.

The overlap with the design-guidance family is deliberate and shallow: this package links to the vendor references for what a model can do and spends its length on the generated artifact and its checks.

## How to read this landscape

Names on this page are leads, not dependencies. A community skill may be unmaintained, disagree with current vendor documentation, or target a model version that no longer exists. Star and install counts are not a selection signal here; what counts is whether a skill produces the artifact in hand and matches the vendor's current page. No vendor other than TypeSafe and OpenRouter published a decision-API skill on 2026-11-08.

## Unverified items

- Install counts and registry listings of community skills: one registry search on one day, not relied on.
- Forks and mirrors of `jev-triage` and `jev-audit`: some repositories are mirrors, not originals; only the originals are listed.
- A skill named like an item above but with unrelated content exists (`hraness/system-one-skills`); match by content, never by name.

## Sources

> Links checked 2026-11-08.

| Claim | Source |
|---|---|
| The vendor skill's repository, release history, and contents | <https://github.com/typesafe-ai/skills> (release v0.6.2, 2026-10-23) |
| The install routes, the one-method rule, and the stale-copy failure | <https://docs.typesafe.ai/agent-skill.md> |
| Community skill locations, contents, and last commits | the repositories named in the table, opened on GitHub on 2026-11-08 |
| The delivery statement and the adopted practices | this package's own rules and references |

### Evidence grade

`VENDOR` for the vendor-published skill section, read from its repository and the vendor page on 2026-11-08. `COMMUNITY` for the community table: each row was opened, but nothing there is endorsed or checked for accuracy. `LOCAL` for the delivery statement.
rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit.Stars: 469kUpdated Claude CodeCodexVerifiedInstall
rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure.Stars: 469kUpdated Claude CodeCodexVerifiedInstall
rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics.Stars: 469kUpdated Claude CodeCodexVerifiedInstall
ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first.Stars: 194kUpdated Claude CodeCodexVerifiedInstall