MantaDigital/stingray-agent-skills / skills/stingray
stingray
VerifiedTrading strategy builder for coding agents. Use when Codex, Claude, Cursor, or a SKILL.md host needs to turn a trading idea into market Evidence, an inspectable Signal, a private Replay, Replay status, or an opt-in Studio Publication. Stingray does not place trades.
| File | Kind | Size |
|---|---|---|
| LICENSE.txt | markdown | 11.1 KB |
| README.md | markdown | 4.3 KB |
| SKILL.md | skill md | 10.8 KB |
| prompts.md | markdown | 3.6 KB |
| references/access-policy.md | markdown | 2.3 KB |
| references/agent-positioning.md | markdown | 2.5 KB |
| references/business-capabilities.md | markdown | 2.8 KB |
| references/capabilities.json | config | 5.8 KB |
| references/co-development.md | markdown | 1.8 KB |
| references/data-coverage.md | markdown | 1.3 KB |
| references/examples.md | markdown | 3.2 KB |
| references/intent-rubrics.md | markdown | 1.9 KB |
| references/north-star-scenarios.md | markdown | 2.4 KB |
| references/replays-and-publications.md | markdown | 3.3 KB |
| references/signal-definitions.md | markdown | 1.8 KB |
| references/token-lifecycle.md | markdown | 1.5 KB |
| references/troubleshooting.md | markdown | 1.6 KB |
| references/workflows.md | markdown | 2.3 KB |
README.md
View on GitHub# Stingray
Human quickstart for the public Stingray skill bundle: a Studio-native market
agent for Codex, Claude Code, Cursor, and other SKILL.md-compatible agents.
## Install
### Recommended: Ask Your Agent
Ask your agent:
```text
Install the skills from MantaDigital/stingray-agent-skills globally for all my coding agents using npx skills.
```
### With `skills.sh` / `npx skills`
```bash
npx skills add MantaDigital/stingray-agent-skills -g -y
```
Install the skill from the repository root, then complete the one-time
credential setup below.
Some agents load skills only when a session starts. If the current session
installed Stingray but cannot use it yet, restart the agent session and ask
again.
## 1. Start With Hello World
The installer may only say the skill was installed. The welcome happens in your
agent.
Paste this into a new agent chat:
```text
Welcome me to Stingray. Check my API token, then run the hello-world thesis: BTC pullback check, when BTCUSDT drops 3% in 24h, replay what happened next over 365 days. Create the Studio Idea and Signal, run the Replay, then publish a public Studio demo link I can open in my browser. Keep live monitoring off.
```
If the token is already set, the agent will create one public auditable Studio
demo and return a browser link. If the token is missing, it will guide setup
without asking you to paste the secret into chat.
## 2. Get a Studio Skills API Token
Studio Skills API tokens are currently provisioned for the private-beta Skills
API surface. Ask your Stingray contact for a Studio Skills API token with the
`skills:full` scope.
The token format starts with `sa_pat_...`.
## 3. Configure the Token
The agent will give you a one-line shell command to paste into your
**terminal** (not the chat). The token never enters the agent's context. Or set
`STINGRAY_PAT` in your shell config for env-var auth.
The skill deliberately does not accept tokens pasted into the agent chat. That
would leak the secret into chat history and the LLM context.
## 4. Start Asking
Once the token is saved, the skill calls the production Studio Skills API:
```text
https://stingray.fi/api/studio/v1
```
Best first prompt:
```text
Welcome me to Stingray. Check my API token, then run the hello-world thesis: BTC pullback check, when BTCUSDT drops 3% in 24h, replay what happened next over 365 days. Create the Studio Idea and Signal, run the Replay, then publish a public Studio demo link I can open in my browser. Keep live monitoring off.
```
The agent will work through:
```text
Idea -> Signal candidate -> accepted Signal -> private Replay -> optional Studio Publication
```
Example prompts:
> "What can Stingray Skills do from this agent?"
> "Use Stingray to test this Hyperliquid thesis privately: ETH funding heat check. Trigger when ETH funding on Hyperliquid rises above 0.75 bps/hr. Replay the last 365 days and report event count, average gap, and whether forward-return samples are available. Do not enable monitoring and do not publish a public Studio link."
> "Confirm that Hyperliquid OI, whale, liquidation, and mark-to-liquidation Signals are not supported in the public skill today, then give me the nearest supported Hyperliquid funding-rate example."
> "Continue the current Idea, answer any missing questions, accept the staged Signal if it looks reasonable, then run a private Replay."
More copy-paste examples live in [`prompts.md`](prompts.md).
## What The Skill Covers
- Studio-native Idea intake and refinement
- Evidence grounding inside a bound Idea
- Signal design through the Studio assistant runtime
- Explicit staged Signal acceptance through `artifact.accept`
- Private Replay through `signal.replay`
- Lightweight status and recovery through `signal.status`, run lookup, and `client_request_id`
- Opt-in public Studio Publication through `idea.publish`
- Hyperliquid boundaries: funding-rate Signals are the current supported demo path; OI, whale, liquidation, and mark-to-liquidation primitives are not supported in the public skill today
- Token isolation, token scope boundary, and privacy-safe debug reporting
## Bundle Contents
```text
skills/stingray/
├── LICENSE.txt
├── prompts.md
├── README.md
├── SKILL.md
└── references/
├── capabilities.json
└── ...
```
## Support
For packaging or distribution issues, contact `dev@mantadigital.io`.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |