Skip to content

junsier9/quant-alpha-research-skill / skills/quant-alpha-research

quant-alpha-research

Verified

Fail-closed crypto quant alpha research for repo-native Stage 0/0.5 validation, falsification, provider concordance, and promotion decisions. Use to find/validate alpha, execute quant research roadmaps, audit promotion evidence, turn market commentary into hypotheses, or decide if alpha reruns are allowed; Chinese triggers include 寻找新alpha、验证这个alpha、执行量化研究路线图 and synonyms. Do not use for generic finance, CSV/charting, product Stage 0, unrelated provider debugging, or GitHub/frontend work.

references/output_templates.md

View on GitHub
# Output Templates

Use these templates only when the user needs a decision-grade report. Keep chat answers shorter when the task is small.

## Alpha decision

```text
decision: no-go | watch | blocked | research-only | go
candidate:
parent_or_control:
horizon:
landing_shape:
data_sources:
data_trust:
stage0_evidence:
falsification_status:
promotion_status:
main_blockers:
next_executable_step:
artifacts_checked:
commands_run:
```

Rules:

- Put the decision first.
- Use `go` only if the repo's promotion gate is actually satisfied.
- Use `blocked` when data trust or pipeline state prevents an honest result.
- Use `research-only` for useful mechanisms that should not touch live or canonical state.

## Provider/data decision

```text
decision: alpha_rerun_allowed = true | false
provider:
coverage_status:
concordance_status:
symbols_or_fields_failed:
closed_bar_handling:
provenance_status:
research_unlocked:
research_blocked:
next_data_step:
```

Rules:

- Never treat coverage as concordance.
- List suspicious symbols or fields explicitly.
- Prefer closed-bar-aware comparisons and symbol-level failures.

## External-source conversion

```text
source_claims:
repo_verified_facts:
mechanism:
candidate_landing_shapes:
minimum_data_needed:
stage0_design:
falsification_tests:
promotion_boundary:
```

Rules:

- Keep source claims separate from repo evidence.
- Translate commentary into a testable state variable or selection rule.
- Name the first falsification test before naming a candidate winner.

## Roadmap execution closeout

```text
roadmap_item:
work_done:
artifacts_created_or_checked:
tests_or_runs:
decision:
remaining_blockers:
next_item:
source_repo_write_status:
```

Rules:

- If the task was read-only, `source_repo_write_status` must state that no source-repo writes were made.
- If a roadmap item is blocked, identify the first unblocker instead of broadening the plan.
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: 471kUpdated 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: 471kUpdated Claude CodeCodexVerifiedInstall
rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics.Stars: 471kUpdated 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