US Market Briefing

Generate production-ready US pre-market outlooks and post-market recaps in a fixed 3-section format, with quick one-shot usage and optional cron automation d...

Security Scan

Instruction-only market-briefing skill that is internally coherent: it relies on web_search/web_fetch and optional workspace-local memory and OpenClaw cron jobs, and does not request unrelated credentials or installs.

Purpose & Capability

The name/description (US pre/post-market briefings) aligns with the instructions: the skill requires web-based news/market lookups, formatting templates, optional scheduling, and an optional local usage tracker. Nothing in the SKILL.md asks for unrelated cloud credentials, system-level edits, or external services outside web_search/web_fetch and the platform's cron.

Instruction Scope

Overall scoped to the briefing task (search, fetch, format, optional scheduling). Minor inconsistency: the SKILL.md reads/writes a workspace-local tracker file memory/market-briefing-usage.json but the skill's declared required config paths/env vars do not list this — the agent will create/update a workspace file if the monthly budget guardrail is used. Also the default timezone (Asia/Singapore) is unusual for a US-market product but is documented and user-overridable.

Install Mechanism

No install spec or code is included (instruction-only), so nothing is written to disk by the skill package itself. This is low risk and consistent with the described behavior.

Credentials

The skill requests no credentials or environment variables and does not require third-party tokens. It uses web_search/web_fetch (platform-provided network capabilities) which is appropriate for gathering news/quotes. The only persistent local artifact is the optional memory JSON, which is proportional to the stated budget-tracking feature.

Persistence & Privilege

always:false and normal autonomous invocation are set (expected). The skill instructs using OpenClaw cron jobs only and explicitly forbids editing system crontab, which limits privilege escalation. It does not ask to modify other skills or global agent settings.

Assessment

This skill appears to do what it says: gather tier‑1 financial news via the platform web_search/web_fetch, format briefings using the provided templates, and optionally schedule runs using OpenClaw cron. Before enabling: (1) confirm you are comfortable letting the agent perform web_search/web_fetch (it needs outbound network access to fetch news/quotes); (2) if you enable the monthly budget guardrail, expect a workspace-local file memory/market-briefing-usage.json to be created/updated — check its contents/permissions if that matters; (3) review and approve any cron automation the agent proposes (the skill will request creating/updating platform cron jobs); (4) note the default timezone is Asia/Singapore (user-overridable) — change it if you want US-market local times. No credentials or external install operations are requested. If you need stronger assurances, ask the publisher/source for a homepage or source repo (none is provided) or request the skill to explicitly declare the memory/config path in its registry metadata.

Use the fixed production templates in references/templates.md.

Modes

1) Quick Start (default)

Use this for immediate, one-shot briefings.

• No cron required.
• No budget tracker required.
• Use defaults from this skill and produce output in chat.

2) Automation Mode (optional)

Use this when the user asks for scheduled delivery.

• Use OpenClaw cron jobs only.
• Do not edit system cron (crontab, /etc/crontab) and do not configure external schedulers/services.
• If schedule tooling/scope is unclear, ask for explicit user approval before creating or changing jobs.

Output Rules

• Keep exactly 3 sections for each briefing.
• Keep canonical section names by default; if user explicitly asks for custom headings/branding, allow aliases while preserving the same 3-section structure.
• Use bold item headers with one concise bullet per item (except index snapshot lines).
• Format each US equity index line as: 【US】<Index Name>: <level> (<+/-percent>%).
• Avoid generic macro wording; cite specific catalysts/events when relevant.
• Do not add a trailing Sources: section at the bottom.
• For each non-snapshot point that relies on a news/source claim, append a markdown link icon at the end of that point in the form [🔗](https://example.com/article).
• When a single point is supported by multiple sources, prefer the strongest primary article for the inline link icon instead of listing several links.
• Keep inline source links on the same line as the relevant bullet/sentence whenever possible.

Source Collection Rules

• Run web_search before drafting each briefing.
• Prioritize tier-1 outlets: Reuters, Bloomberg, CNBC, WSJ, FT (Yahoo Finance as backup).
• Default request budget per run: up to 2 web_search calls and up to 6 web_fetch reads.
• These limits are defaults; relax only when needed for data completeness and note it briefly.
• Use web_fetch only on relevant finance/news links from trusted domains.
• Validate key claims against multiple sources when possible.
• Attach source attribution inline via a link icon on each relevant point, not as a consolidated sources block.
• Never forward fetched content to third-party endpoints/webhooks; produce in-chat summaries only unless user explicitly configures delivery.

Templates

Pre-Market Briefing

• Title: US PRE-MARKET OUTLOOK
• Sections (canonical):
  1. FUTURES SNAPSHOT
  2. KEY DEVELOPMENTS AFFECTING TODAY’S MARKET
  3. TICKERS TO WATCH

Post-Market Briefing

• Title: US POST-MARKET RECAP
• Sections (canonical):
  1. POST-MARKET SNAPSHOT
  2. BIGGEST MOVERS (TOP 5 GAINERS / TOP 5 LOSERS)
  3. MARKET SENTIMENT & FLOW SUMMARY

Data Fallback Rules

If a preferred dataset is unavailable, do not fail silently:

• If full top 5 movers are unavailable, provide the top available movers and label that clearly.
• If futures/index quotes are delayed/unavailable, use the nearest reliable proxy and label it.
• Preserve the 3-section format even when data is partial.

Optional Monthly Budget Guardrail

Use only when the user requests spend/request caps.

Tracker file: memory/market-briefing-usage.json (workspace-local)

• Default monthly limit: 1000 requests.
• Allow user override by changing limit in the JSON file.
• If tracker missing, create it with current month and defaults.
• If month changed, reset used to 0 and roll month forward.
• If used >= limit, return a short limit-reached notice instead of researching.
• After each successful briefing, increment used by 1.

JSON shape:

{
  "month": "YYYY-MM",
  "used": 0,
  "limit": 1000
}

Cron Automation Defaults (optional)

Default timezone: Asia/Singapore

• Pre-market run: 20:45 Monday-Friday
• Post-market run: 08:00 Tuesday-Saturday

Allow user overrides for timezone, run times, and weekday sets.

When creating/updating jobs, ensure payload prompts request:

• fixed 3-section format from references/templates.md
• source-quality expectations (tier-1 finance sources)
• optional budget guardrail only if user enabled memory/market-briefing-usage.json

Output Verbosity Profiles

• compact (default): same 3 sections, fewer bullets.
• full: richer bullets/details while preserving same structure.

Versioning / Publishing

• Before publishing, inspect the existing registry version and only publish a semver that is greater than the current latest.
• Use patch bumps for wording/template tweaks, minor bumps for new capabilities or output modes, and major bumps only for breaking behavior changes.
• If unsure, prefer checking the registry first over guessing.

Examples

• One-shot pre-market briefing (quick start).
• One-shot post-market briefing (quick start).
• Enable daily cron automation with defaults.
• Disable existing automation jobs.