agentleFS
bobmmaobobmmao/agent-workflow-skills/skills/research/us-stock-radar/SKILL.md

us-stock-radar

Skill0 starsChanged 5 months agoMITView on GitHub

Professional US stock radar for screening, deep dives, and watchlist alerts using public market data. Use when the user wants ranked stock candidates, A/B/C/D signal grading, timestamped evidence, confidence-aware summaries, and cleaner pro or beginner explanations for US equities.

The file

Show its rules and text

We split this skill into the rules it gives an agent

15 rules. About 0 are common, 9 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run a practical US stock workflow in 3 modes:

    line 8
  2. Very specific:

    Run scripts/us_stock_radar.py with the appropriate mode.

    line 20
  3. Distinctive:

    Read JSON output first; treat it as the source of truth.

    line 21
  4. Distinctive:

    Avoid deterministic predictions.

    line 23
  5. Distinctive:

    If some endpoints fail, continue with degraded coverage and expose the reduction in confidence.

    line 24
  6. Very specific:

    Verify that availability, data_gaps, and degraded_mode are exposed when coverage is partial.

    line 33
  7. Distinctive:

    Read-only skill: query public market endpoints only.

    line 49
  8. Distinctive:

    Use no authentication, cookies, brokerage accounts, or private APIs.

    line 50
  9. Distinctive:

    Place no orders, execute no trades, and mutate no portfolio state.

    line 51
  10. Distinctive:

    Write no files and send no outbound messages as part of normal use.

    line 52
  11. Distinctive:

    Do not connect a brokerage account unless the user explicitly asks for broker integration in a separate request.

    line 55
  12. Very specific:

    Default language behavior: auto.

    line 66
  13. Very specific:

    If --lang auto and the prompt contains Chinese, switch final narrative to Chinese.

    line 67
  14. Very specific:

    If --lang auto and no Chinese is detected, use English.

    line 68
  15. Very specific:

    If availability.sec_fundamentals is true, reported fundamentals were filled from SEC Company Facts.

    line 111

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

US Stock Radar

Run a practical US stock workflow in 3 modes:

  • screener: rank a ticker universe by multi-factor signal score
  • deep-dive: analyze one ticker with fundamentals + technical proxies
  • watchlist: monitor custom tickers and output alert candidates
  • daily purchase review: run the OpenClaw wrapper that combines screening, market context, runtime checks, and a user-confirmation prompt

This skill is a read-only heuristic market workflow, not a full institutional research terminal. Public free endpoints may be partial, delayed, or rate-limited; surface those gaps explicitly.

Reported fundamentals are filled from SEC EDGAR Company Facts when Yahoo quote/summary endpoints are intentionally disabled. The SEC fallback uses local caching and serial throttling; it is suitable for PE/PB/ROE/revenue-growth style reported metrics, not forward PE or analyst estimates.

Workflow

  1. Run scripts/us_stock_radar.py with the appropriate mode.
  2. Read JSON output first; treat it as the source of truth.
  3. Explain conclusions with explicit caveats, confidence, and data gaps.
  4. Avoid deterministic predictions; present signal grade, trigger reasons, and partial-data warnings.
  5. If some endpoints fail, continue with degraded coverage and expose the reduction in confidence.

Quick Audit Path

For a fast review:

  1. Run python3 skills/us-stock-radar/scripts/us_stock_radar.py --sources
  2. Run python3 skills/us-stock-radar/scripts/us_stock_radar.py --mode screener --json
  3. Confirm the script only performs read-only public HTTP requests.
  4. Verify that availability, data_gaps, and degraded_mode are exposed when coverage is partial.

Commands

python3 skills/research/us-stock-radar/scripts/us_stock_radar.py --sources
python3 skills/research/us-stock-radar/scripts/us_stock_radar.py --version
python3 skills/research/us-stock-radar/scripts/us_stock_radar.py --mode screener --tickers "AAPL,MSFT,NVDA,AMZN,GOOGL"
python3 skills/research/us-stock-radar/scripts/us_stock_radar.py --mode deep-dive --ticker AAPL --audience pro
python3 skills/research/us-stock-radar/scripts/us_stock_radar.py --mode deep-dive --ticker TSLA --audience beginner --lang zh
python3 skills/research/us-stock-radar/scripts/us_stock_radar.py --mode watchlist --tickers "AAPL,NVDA,TSLA" --event-mode high-alert
python3 skills/research/us-stock-radar/scripts/us_stock_radar.py --mode screener --json

Safety / Scope Boundary

  • Read-only skill: query public market endpoints only.
  • Use no authentication, cookies, brokerage accounts, or private APIs.
  • Place no orders, execute no trades, and mutate no portfolio state.
  • Write no files and send no outbound messages as part of normal use.
  • Produce analysis only; not investment advice.
  • Daily purchase review must ask for explicit user confirmation before any real-world action.
  • Do not connect a brokerage account unless the user explicitly asks for broker integration in a separate request.

Optional Local Integration

  • Runtime: configure a Python virtual environment with the dependencies required by the script.
  • Config: keep local config files outside the repository when they contain personal workflow settings.
  • Logs: write runtime logs to a local ignored directory.
  • Reference repos: keep external research repos outside this repository unless they are vendored intentionally.

Output Policy

  • Default language behavior: auto.
  • If --lang auto and the prompt contains Chinese, switch final narrative to Chinese.
  • If --lang auto and no Chinese is detected, use English.
  • --json output is language-neutral.
  • Always include:
    • as_of_utc
    • mode
    • event_mode
    • availability
    • data_gaps
    • degraded_mode
    • confidence
    • sources
    • heuristic notes / caveats
  • Audience modes:
    • pro: concise signal summary
    • beginner: plain-language interpretation
  • Event modes:
    • normal
    • high-alert (stricter thresholds)

Scoring (A/B/C/D)

Signal score combines heuristic checks such as:

  • valuation range (PE)
  • RSI health
  • volume expansion
  • price vs MA50
  • revenue growth
  • ROE quality

Grades:

  • A: score >= 5
  • B: score = 4
  • C: score = 2-3
  • D: score <= 1

Interpretation Guardrails

  • Grades are heuristic summaries, not price targets.
  • Missing fundamentals should lower confidence rather than silently default bullish/bearish.
  • Free-market endpoints can be delayed, partially populated, or blocked by region.
  • Premarket / scheduled workflows should keep timestamps explicit.
  • In OpenClaw/Bounty conversations, do not fetch Yahoo chart URLs through web_fetch and do not paste raw Yahoo JSON. Use the local script and summarize its JSON output.
  • In OpenClaw/Bounty conversations, prefer --compact-json. Do not describe *_yahoo_quote_skipped or *_yahoo_summary_skipped as current Yahoo rate limiting; those are intentional protection settings unless actual 429/cooldown signals are present.
  • If availability.sec_fundamentals is true, reported fundamentals were filled from SEC Company Facts. Do not call fundamentals missing just because forward_pe is still unavailable.

Data Sources

  • Yahoo Finance quote API: /v7/finance/quote
  • Yahoo Finance chart API: /v8/finance/chart
  • Yahoo Finance quoteSummary API: /v10/finance/quoteSummary
  • SEC EDGAR Company Facts API: /api/xbrl/companyfacts/CIK##########.json
  • Stooq public fallback: daily quote / history CSV endpoints

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.