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

portfolio-watch

Skill0 starsChanged 4 months agoMITView on GitHub

Read the user's holdings (via a read-only broker MCP if connected, else a manual-paste snapshot) and flag any that need action today — stop hit, thesis/rating break, regime flip, or a name that entered ASM/GSM. Use for daily portfolio review and scheduled monitoring. Broker-agnostic, read-only; never places orders.

3 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 3 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read holdings through the india-data port, in trust order:

    line 9
  2. Very specific:

    If neither is available (e.g.

    line 12
  3. Distinctive:

    Stop hit / trailing-stop breach (swing: structure/ATR trail.

    line 20
  4. Distinctive:

    Never square off or place any order — output is an alert for the human to act on in their broker app.

    line 27
  5. Distinctive:

    Never report a position or price you did not read this run (from the connector or the paste).

    line 28

“Common” is measured: kept by at least 6 in 10 of the 3 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

Portfolio Watch (hold / trim / exit)

Get holdings — broker-agnostic, fail-closed

Read holdings through the india-data port, in trust order:

  1. A read-only broker MCP (Upstox/Kotak preferred; if Zerodha/Groww, reads only — the write tools are deny-hooked), when a session is authenticated and fresh → delay_class: live.
  2. Else a manual-paste snapshot (SYMBOL QTY AVG_COST) → delay_class: manual. If neither is available (e.g. logged-out broker session and no paste), say so and STAND DOWN — never guess positions or prices. Stamp every reading {value, source, as_of, delay_class}.

Run the deterministic gate — capability resolver

Gate MCP → python3 "${CLAUDE_PLUGIN_ROOT}/scripts/kit.py" → JS-paise Artifact → REFUSE. Never compute the numbers in prose.

Procedure

  1. For each holding determine the tier (swing vs positional) from how it was entered (or ask once).
  2. Exit checks:
    • Stop hit / trailing-stop breach (swing: structure/ATR trail; positional: no ATR trail — governance events only).
    • Re-run the exclusion filter: a held name now in ASM/GSM/ESM/T2T or an F&O ban is an exit/caution signal.
    • Regime flip to risk-off → flag the weakest / lowest-RS names first (swing only).
    • Fundamental/governance break (pledge jump, auditor exit, rating downgrade) from india-data/screener.
  3. Report. One concise line per holding: HOLD / TRIM / EXIT + the single dominant reason + the freshness stamp (source, delay_class). Summarise "N holdings, X need action." Where the surface supports it, render a beautiful portfolio Artifact (per-name status + portfolio heat/health); otherwise a text digest. Append the disclaimer.

Never

  • Never square off or place any order — output is an alert for the human to act on in their broker app. Read-only.
  • Never report a position or price you did not read this run (from the connector or the paste).

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.