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

autopilot

Skill0 starsChanged 4 months agoMITView on GitHub

Run the full survival-first review UNATTENDED, end to end — regime check, portfolio hold/trim/exit, and an opportunity scan — then send ONE concise notification. Built for Cloud Routines (scheduled, zero human interaction). Broker-free (free/delayed public data + your last synced holdings snapshot); read-only; never places, modifies, or cancels an order.

82 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use review-bugbot, ultraqa 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.
  • 0mlkrizzz655335vSkill

    review-bugbot

    缺陷导向代码审查。用 Bugbot 子代理找 bug。

    Also in 3 projects that use this

  • Jurel89Skill

    ultraqa

    QA cycling loop — build/lint/test, fix, retry up to 5 cycles

    Also in 3 projects that use this

  • 0mlkrizzz655335vSkill

    rename-chat

    重命名当前聊天。按聊天内容改标题。"disable-model-invocation: true

    Also in 3 projects that use this

  • Tsumugii24Skill

    split-to-prs

    Split current work into small reviewable PRs. Use when the user asks to split a chat, set of changes, branch, or PR.

    Also in 4 projects that use this

  • PandelisZSkill

    update-cli-config

    View and modify Cursor CLI configuration settings in ~/.cursor/cli-config.json. Use when the user wants to change CLI settings, configure permissions, switch approval mode, enable vim mode, toggle display options, configure sandbox, or manage any CLI preferences.

    Also in 3 projects that use this

  • klauswgSkill

    deep-interview

    在明确批准执行之前,通过数学化歧义门控进行苏格拉底式深度访谈

    Also in 5 projects that use this

  • +1 more from 0mlkrizzz655335v

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 73 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. Very specific:

    No fabrication: Every number comes from either (a) the deterministic gate via the capability resolver, or (b) a data read you can cite …

    line 12
  2. Distinctive:

    Never place / modify / cancel an order, or attach a write-capable broker connector.

    line 30
  3. Distinctive:

    Never attempt an interactive broker login in an unattended run.

    line 31
  4. Distinctive:

    Never report a holding / price / status you did not read this run.

    line 32

“Common” is measured: kept by at least 6 in 10 of the 73 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

Autopilot — autonomous survival-first review (read-only, broker-free)

You are running UNATTENDED (a Cloud Routine, or an explicit /autopilot run). There is no human to ask mid-run and no interactive broker login is possible — Indian broker OAuth sessions expire daily (~03:00 IST) and an unattended run cannot re-authenticate. So this run is broker-free by design. Default posture is no-trade / STAND DOWN. Finish with exactly ONE notification.

Hard safety invariants — never violate

  • READ-ONLY, and no broker connector attached. Do not call any order / modify / cancel / GTT tool, ever (a PreToolUse deny hook is the backstop, not your licence to try). Run on free/delayed public data (via india-data: screener + Yahoo, delayed) + the user's last synced holdings snapshot (manual paste, or a prior authenticated sync). If live private state is needed and unavailable → STAND DOWN on that item and say so.
  • No fabrication. Every number comes from either (a) the deterministic gate via the capability resolver, or (b) a data read you can cite {source, as_of, delay_class}. Never invent a price, ATR, level, or surveillance status.
  • Missing/stale → STAND DOWN. Bias to no-trade and name what was missing.
  • Never widen a stop, average down, or override a hard veto.

Run the deterministic gate — capability resolver

  1. A Gate MCP connected → call it. 2. Else code execution (this Routine has a shell) → python3 "${CLAUDE_PLUGIN_ROOT}/scripts/kit.py". 3. Else REFUSE → send a no-trade advisory (there is no Artifact surface in an unattended run). Never compute the gate numbers in prose.

Procedure — run in order, then notify once

  1. Regime — "trade at all today?" Via india-data (delayed public data): index trend (Nifty vs 50/200 DMA), breadth, India VIX. Set posture full / caution / cash; risk-off (Nifty < 200 DMA) → bias hard against new longs. Make "raise cash / stand down" an explicit output when the regime fails — not merely the absence of a buy.
  2. Portfolio review — from the last synced holdings snapshot only (do NOT attempt a live broker read). For each holding: re-run the surveillance/exclusion filter, check stop / thesis / regime, pass through the gate → HOLD / TRIM / EXIT + the single dominant reason. If no snapshot is available, say so and skip (never guess positions).
  3. Opportunity scan. Universe = config/watchlist.json else holdings' sectors else Nifty-500 leaders. Exclusion kill-filter FIRST, then momentum/trend + fundamental red-flags, then the gate per survivor. Keep only decision == long, shares > 0, with a ≥2R target. Rank; take the top 3.
  4. Notify — ONE concise message:
    • Regime: full / caution / cash + reason (and any raise-cash call).
    • Holdings needing action: N — one line each (HOLD/TRIM/EXIT + reason), or "none".
    • Top opportunities: up to 3, each symbol · entry · stop · size · target(≥2R) · net break-even · one reason. If none: "No trade — dominant gate: <what eliminated the most names>."
    • The disclaimer.

Never

  • Never place / modify / cancel an order, or attach a write-capable broker connector. Output is an alert; the human acts in their own broker.
  • Never attempt an interactive broker login in an unattended run.
  • Never report a holding / price / status you did not read this run. Keep the notification short — it is the whole payload.

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.