Index of every page on this site: llms.txt
All skills
tradermonty avatar

/edge-candidate-agent

@f95e770
Run it yourself

Generate and prioritize US equity long-side edge research tickets from EOD observations, then export pipeline-ready candidate specs for trade-strategy-pipeline Phase I. Use when users ask to turn hypotheses/anomalies into reproducible research tickets, convert validated ideas into `strategy.yaml` + `metadata.json`, or preflight-check interface compatibility (`edge-finder-candidate/v1`) before running pipeline backtests.

Use this Skill: https://skilld.dev/gh/tradermonty/claude-trading-skills/edge-candidate-agent

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

SKILL.md

≈111 tokens for metadata: the name and description. ≈1.3k when used: this file. ≈1.4k more on demand in 5 files.

Description uses 5.6% of example budget

Before choosing a Skill, your Agent reads its name and description. All available Skills share that space.

  • A shorter description leaves more room for other Skills. This entry exceeds our 1% size suggestion.

In our Claude Code example, all Skill names and descriptions share 8,000 characters. This Skill uses ≈446 characters, or 5.6%.

The 1% threshold is a size suggestion. Longer descriptions can still fit.

Your model, settings, and other Skills decide how much text your Agent can read.

Example settings and source

The example uses a 200k-token context and default Claude Code settings. The count includes the name, description, separators, and when_to_use when present. Codex also counts local file paths.

Skit's source and limits: Codex 0.160.1, Claude Code 2.1.292.

Edge Candidate Agent

Overview

Convert daily market observations into reproducible research tickets and Phase I-compatible candidate specs. Prioritize signal quality and interface compatibility over aggressive strategy proliferation. This skill can run end-to-end standalone, but in the split workflow it primarily serves the final export/validation stage.

When to Use

  • Convert market observations, anomalies, or hypotheses into structured research tickets.
  • Run daily auto-detection to discover new edge candidates from EOD OHLCV and optional hints.
  • Export validated tickets as strategy.yaml + metadata.json for trade-strategy-pipeline Phase I.
  • Run preflight compatibility checks for edge-finder-candidate/v1 before pipeline execution.

Prerequisites

  • Python 3.9+ with PyYAML installed.
  • Access to the target trade-strategy-pipeline repository for schema/stage validation.
  • uv available when running pipeline-managed validation via --pipeline-root.

Output

  • strategies/<candidate_id>/strategy.yaml: Phase I-compatible strategy spec.
  • strategies/<candidate_id>/metadata.json: provenance metadata including interface version and ticket context.
  • Validation status from scripts/validate_candidate.py (pass/fail + reasons).
  • Daily detection artifacts:
    • daily_report.md
    • market_summary.json
    • anomalies.json
    • watchlist.csv
    • tickets/exportable/*.yaml
    • tickets/research_only/*.yaml

Position in Split Workflow

Recommended split workflow:

  1. skills/edge-hint-extractor: observations/news -> hints.yaml
  2. skills/edge-concept-synthesizer: tickets/hints -> edge_concepts.yaml
  3. skills/edge-strategy-designer: concepts -> strategy_drafts + exportable ticket YAML
  4. skills/edge-candidate-agent (this skill): export + validate for pipeline handoff

Workflow

  1. Run auto-detection from EOD OHLCV:
    • skills/edge-candidate-agent/scripts/auto_detect_candidates.py
    • Optional: --hints for human ideation input
    • Optional: --llm-ideas-cmd for external LLM ideation loop
  2. Load the contract and mapping references:
    • references/pipeline_if_v1.md
    • references/signal_mapping.md
    • references/research_ticket_schema.md
    • references/ideation_loop.md
  3. Build or update a research ticket using references/research_ticket_schema.md.
  4. Export candidate artifacts with skills/edge-candidate-agent/scripts/export_candidate.py.
  5. Validate interface and Phase I constraints with skills/edge-candidate-agent/scripts/validate_candidate.py.
  6. Hand off candidate directory to trade-strategy-pipeline and run dry-run first.

Quick Commands

Daily auto-detection (with optional export/validation):

python3 skills/edge-candidate-agent/scripts/auto_detect_candidates.py \
  --ohlcv /path/to/ohlcv.parquet \
  --output-dir reports/edge_candidate_auto \
  --top-n 10 \
  --hints path/to/hints.yaml \
  --export-strategies-dir /path/to/trade-strategy-pipeline/strategies \
  --pipeline-root /path/to/trade-strategy-pipeline

Create a candidate directory from a ticket:

python3 skills/edge-candidate-agent/scripts/export_candidate.py \
  --ticket path/to/ticket.yaml \
  --strategies-dir /path/to/trade-strategy-pipeline/strategies

Validate interface contract only:

python3 skills/edge-candidate-agent/scripts/validate_candidate.py \
  --strategy /path/to/trade-strategy-pipeline/strategies/my_candidate_v1/strategy.yaml

Validate both interface contract and pipeline schema/stage rules:

python3 skills/edge-candidate-agent/scripts/validate_candidate.py \
  --strategy /path/to/trade-strategy-pipeline/strategies/my_candidate_v1/strategy.yaml \
  --pipeline-root /path/to/trade-strategy-pipeline \
  --stage phase1

Export Rules

  • Keep validation.method: full_sample.
  • Keep validation.oos_ratio omitted or null.
  • Export only supported entry families for v1:
    • pivot_breakout with vcp_detection
    • gap_up_continuation with gap_up_detection
  • Mark unsupported hypothesis families as research-only in ticket notes, not as export candidates.

Guardrails

  • Reject candidates that violate schema bounds (risk, exits, empty conditions).
  • Reject candidate when folder name and id mismatch.
  • Require deterministic metadata with interface_version: edge-finder-candidate/v1.
  • Use --dry-run in pipeline before full execution.

Resources

skills/edge-candidate-agent/scripts/export_candidate.py

Generate strategies/<candidate_id>/strategy.yaml and metadata.json from a research ticket YAML.

skills/edge-candidate-agent/scripts/validate_candidate.py

Run interface checks and optional StrategySpec/validate_spec checks against trade-strategy-pipeline.

skills/edge-candidate-agent/scripts/auto_detect_candidates.py

Auto-detect edge ideas from EOD OHLCV, generate exportable/research tickets, and optionally export/validate automatically.

references/pipeline_if_v1.md

Condensed integration contract for edge-finder-candidate/v1.

references/signal_mapping.md

Map hypothesis families to currently exportable signal families.

references/research_ticket_schema.md

Ticket schema used by export_candidate.py.

references/ideation_loop.md

Hint schema and external LLM ideation command contract.

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

1 alert6d5 checks · Risk MEDIUM
  • Gen Agent Trust Hub6d

    The skill provides scripts that can execute arbitrary external commands and code from user-specified directories, posing a risk of command injection or remote code execution if the agent is manipulated.

  • Socket6d

    1 alert: gptAnomaly

  • Snyk6d

    Risk: LOW · No issues

  • Runlayer8mo

    5/15 files flagged

  • ZeroLeaks7mo

    Score: 93/100 · 2 sections analyzed

Signed by skilld at f95e770. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub yesterday.

Activeupdated 9 months ago

README badge

README badge for tradermonty/claude-trading-skills/edge-candidate-agent