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 |
references/north-star-scenarios.md
View on GitHub# North-Star Scenarios
These are multi-step product loops external agents should be able to complete
through the public skill.
## 1. Public Demo From A Generic Thesis
Goal: create a public artifact that demonstrates Stingray without exposing
private user data.
Flow:
1. `idea.intake`
2. `signal.design`
3. `artifact.accept`
4. `signal.replay`
5. `signal.status` or lookup
6. `idea.publish`
Use a generic thesis such as BTC pullback. Keep live monitoring off.
## 2. Private Research Loop
Goal: help the user reason about a thesis before sharing or activating anything.
Flow:
1. `idea.intake`
2. `evidence.ground`
3. `signal.design`
4. ask questions if `needs_input`
5. stop at staged candidate unless the user wants acceptance
## 3. Closed Signal Replay Loop
Goal: start from a raw idea, create a committed Signal, and privately Replay it.
Flow:
1. `idea.intake`
2. `signal.design`
3. `artifact.accept`
4. `signal.replay`
5. `signal.status` or lookup routes
6. summarize the Replay and keep it private
## 4. Hyperliquid Funding Replay
Goal: show a concrete Hyperliquid workflow that is replayable today.
Flow:
1. `idea.intake`
2. `evidence.ground` to confirm funding coverage
3. `signal.design`
4. `artifact.accept`
5. `signal.replay`
6. summarize event count, clustering/gap, and forward-return availability when
returned
## 5. Unsupported Hyperliquid Primitive
Goal: handle OI, whale, liquidation, or mark-to-liquidation requests honestly
without implying current support.
Flow:
1. explain from the skill docs that the primitive is not supported in the public
skill today
2. offer a funding-rate Signal example or a privacy-safe product feedback note
3. do not call `idea.intake`, `evidence.ground`, or `signal.design` merely to
test the unsupported primitive
## 6. External-Agent Recovery
Goal: recover from timeout, lost context, or long-running Replay.
Flow:
1. use `run_id` lookup when available;
2. use `client_request_id` lookup when only caller recovery key is available;
3. use `signal.status` when Idea and Signal ids are known;
4. retry only if recovery shows no active or completed run.
## 7. Publication After Review
Goal: publish only after the user has inspected the private result.
Flow:
1. summarize the private Replay;
2. ask whether the user wants a public Studio Publication if they have not
already asked;
3. call `idea.publish` only after explicit consent.
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: 323k | 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: 323k | 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: 323k | Updated | Claude CodeCodex | Verified | Install |