agentleFS
sumeetkbhardwajsumeetkbhardwaj/small-trading-indian-stock-market-survival-kit/skills/check/SKILL.md

check

Skill0 starsChanged 4 months agoMITView on GitHub

Run the full survival-first gate stack on ONE stock the user is tempted to buy right now, and return a plain verdict (TRADE / NO-TRADE) with the single dominant reason and — if rejected — the counterfactual cost it must clear anyway. Works on any surface via the capability resolver; broker-agnostic data. Use when the user asks "should I buy X", "what about X", "check X", or names a single NSE/BSE ticker. Read-only, advisory only.

88 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use daily-arxiv, exp-eval and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • LambenthanSkill

    daily-arxiv

    每日从 arXiv 拉取新论文,过滤相关性,auto-ingest 高优先级论文,检测 SOTA 更新

    Also in 5 projects that use this

  • LambenthanSkill

    exp-eval

    实验判决门:Review LLM 独立评判实验结果 → 4 种判决路径 → 自动更新 claims confidence、ideas status、graph edges

    Also in 5 projects that use this

  • skyllwtSkill

    exp-pilot-eval

    Pilot result evaluation — read pilot results, apply success criteria, update idea page (pilot_result, failure_reason if failed), generate PILOT_VERDICT_REPORT. Called by /ideate Phase 5 after /exp-pilot-run.

    Also in 4 projects that use this

  • skyllwtSkill

    exp-pilot-run

    Pilot experiment execution — read Pilot Spec YAML, write pilot code, run experiment(Confirm with the user before operation and require the applicant to conduct manual inspection), return results. Called by /ideate Phase 5. Does NOT modify wiki pages or judge pass/fail.

    Also in 4 projects that use this

  • JuliusBrusseeSkill

    backprop

    Bug → spec protocol. When a bug is found or a test fails, trace the cause, decide whether a new §V invariant would catch recurrence, append to §B. This is the one non-obvious thing SDD does that plan-then-execute doesn't. Triggers on test failure, bug report, post-mortem, or explicit user ask.

    Also in 4 projects that use this

  • +6 more from Lambenthan
  • +1 more from skyllwt

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 3 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 75 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Always end with the disclaimer (present in the gate output).

    line 25
  2. Distinctive:

    Never override a hard veto (surveillance, stale data, negative net edge, extended past pivot, target < 2R, overtrading) because the user …

    line 28
  3. Distinctive:

    Never place or suggest placing an order.

    line 29
  4. Very specific:

    Never report a price/ATR/fundamental you did not fetch this run, and never compute the gate numbers yourself.

    line 30

“Common” is measured: kept by at least 6 in 10 of the 75 edited versions of this file we could read. “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

Single-name check (meet the impulse, then teach)

The user named ONE ticker they are tempted by. Do not lecture. Run the gates and give a straight answer, fast. Default is NO-TRADE.

Run the deterministic gate — capability resolver (never compute the numbers in prose)

Pick the highest-fidelity path available:

  1. A Gate MCP tool is connected → call it.
  2. Else code execution is available (Claude Code / Cloud Routine / paid-tier chat container) → python3 "${CLAUDE_PLUGIN_ROOT}/scripts/kit.py" evaluate (stdin = candidate JSON).
  3. Else plain chat → hand the user the JS-paise gate Artifact (client-side, model out of the loop) and read the verdict off it.
  4. Else REFUSE: say plainly the gate could not run, give an advisory-only read, and default to NO-TRADE.

Never assert a sizing / cost / break-even / R number the gate did not compute.

Procedure

  1. Fetch the ticker's data via the india-data skill (broker-agnostic: free-web delayed [screener + Yahoo] + manual-paste baseline; a read-only broker MCP is optional enrichment). Stamp every value {value, source, as_of, delay_class} and label delayed data as delayed. Need: last close/quote, ATR, 200-DMA/trend, surveillance status (ASM/GSM/ESM/T2T/circuit), liquidity, obvious fundamental red flags, and a target (for the 2R gate). If required data is missing or stale → STAND DOWN, say what was missing, never guess a price.
  2. Build the evaluate JSON — entry, stop (structure/ATR), targets (drives the exit contract + 2R gate), equity/risk/cap/heat/gross/exposure from config/, and freshness reflecting the reading's delay_class. Run the gate via the resolver above.
  3. Present, plainly:
    • TRADE → "Passes. Entry X, stop Y, size N shares, target Z (≥2R), break-even B%." Then the one reason it's worth it, plus the exit contract (stop · R-target · trailing rule) from the gate output.
    • NO-TRADE → "No. Dominant reason: <the veto>." Then the counterfactual cost — the round-trip break-even it must clear even if it worked. That number is the teaching, not a scold.
  4. Always end with the disclaimer (present in the gate output).

Never

  • Never override a hard veto (surveillance, stale data, negative net edge, extended past pivot, target < 2R, overtrading) because the user really wants in. State it; the decision is theirs in their own broker.
  • Never place or suggest placing an order. Read-only.
  • Never report a price/ATR/fundamental you did not fetch this run, and never compute the gate numbers yourself.

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.