Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add Haruspex-guru/haruspex-skills --skill haruspex-watchlist-reviewgit clone --depth 1 https://github.com/Haruspex-guru/haruspex-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/haruspex-guru/haruspex-skills/haruspex-watchlist-review)<a href="https://agentmods.dev/skills/haruspex-guru/haruspex-skills/haruspex-watchlist-review"><img src="https://agentmods.dev/badge/skills/haruspex-guru/haruspex-skills/haruspex-watchlist-review/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/haruspex-guru/haruspex-skills/haruspex-watchlist-review"><img src="https://agentmods.dev/badge/skills/haruspex-guru/haruspex-skills/haruspex-watchlist-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00135 | $0.01761 |
| Opus 5 | $0.00068 | $0.00881 |
| Sonnet 5 | $0.00027 | $0.00352 |
| Haiku 4.5 | $0.00014 | $0.00176 |
Grade A, and why
haruspex-watchlist-review scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 160 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Haruspex Watchlist Review
What this skill does
Runs a single batched call against the Haruspex API and produces a scannable table review of a watchlist: ranked by score, grouped by signal, with the biggest movers and the largest dimensional flags surfaced separately.
Optimized for the "morning coffee" use case — a fast read of the user's whole watchlist that surfaces what changed and what to look at more closely.
When to use this skill
Trigger on any of these patterns:
- "Watchlist update" / "Run my watchlist" / "Daily check"
- "How are my stocks doing?"
- The user lists 3 or more tickers in a single message
- "Compare AAPL, MSFT, GOOGL, META, AMZN"
- "Portfolio review" / "Weekly review"
- "Which of these is strongest?" with a list
- "Rank these tickers"
Defer to a sister skill if:
- Only 1–2 tickers mentioned → use
haruspex-stock-analystinstead. - The user states a thesis on a position → use
haruspex-thesis-tracker. - The user is writing in Japanese → use
haruspex-stock-analyst-ja.
Prerequisites
This skill requires @haruspex-guru/mcp-server to be installed and connected.
First action: verify the MCP tools are available.
If get_batch_scores is not available, output the install instructions from
shared/MCP_SETUP.md and stop.
Workflow
Step 0 — Telemetry ping (do this first, always). Before any other tool
call, invoke record_skill_invocation once with skill: "haruspex-watchlist-review"
and client set to the best-guess client surface (e.g. claude-code,
claude-desktop, claude-ai). This is anonymous aggregate telemetry — only
the skill name, version, and client tag travel. No tickers, no conversation,
no user identifiers. The call is fail-silent: if it errors or the tool is
unavailable, ignore it and continue. Do not retry, do not surface the result
to the user, and never block the review on this step.
- Collect the ticker list. If the user provided tickers, parse them. If the user said "my watchlist" without specifying, ask for the list — never assume or invent one. Trim to ≤50 tickers (the batch limit).
- Single batched call. Call
get_batch_scores(symbols=[...]). Do not callget_stock_scoreper ticker — the batched call exists for this. - Handle partial NOT_FOUND. Some tickers in the list may not be in the
Haruspex universe. They will be absent from
data.scores. Compute the missing set and report it honestly. Do not fabricate scores for missing tickers. - Sort and group. Sort returned scores by
scoredescending. Group bysignal(buy / hold / sell). Within each group, list largest score first. - Surface movers. Identify the top 3 by absolute
|change|. These get their own callout — the user wants to know what moved, regardless of direction. - Surface flags. Identify any ticker with a topic dimension worse than
score < 35ANDchange ≤ -5. These are the "watch closely" flags. - Output in the format below.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 160 lines · 135 tokens per session scan A dd95d3dab7ff
haruspex-watchlist-review is a skill published in the GitHub repository Haruspex-guru/haruspex-skills (1 stars, last pushed 22d ago), licensed MIT. It adds 135 tokens to every session and 1,761 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
ib-pmcc-advisor
Analyze PMCC (Poor Man's Covered Call / diagonal spread) positions from IB portfolio. For each diagonal spread, reports short leg risk (delta, IV, assignment probability), daily P&L projections, top-3 roll candidates, and a side-by-side comparison table. Requires TWS or IB Gateway running locally.
scanner-pmcc
Scan stocks for Poor Man's Covered Call (PMCC) suitability. Analyzes LEAPS and short call options for delta, liquidity, spread, IV, yield, trend direction, and earnings proximity. Use when user asks about PMCC candidates, diagonal spreads, or LEAPS strategies.
ib-stop-loss
Downside stop-loss management for PMCC, naked LEAPS, and stock positions in IB. Computes stop prices, detects alerts, and places conditional combo orders. Dry-run by default. Requires TWS or IB Gateway running locally.
ib-trailing-stop
Server-side trailing stop management for stocks and naked LEAPS in IB. Places native TRAIL orders that auto-ratchet the stop as price climbs. Dry-run by default. Requires TWS or IB Gateway running locally.
ib-collar
Generate tactical collar strategy reports for protecting PMCC positions through earnings or high-risk events. Requires TWS or IB Gateway running locally.
ib-find-short-roll
Find roll options for existing short positions OR find best covered call/put to open against long stock. Use when user asks about rolling shorts, finding roll candidates, writing covered calls, or managing option positions. Requires TWS or IB Gateway running locally.