Earnings Review

Event-driven earnings reviews built on primary filings, with quality math and promise tracking computed by code, not prose.

⚠️ **Trading warning / Avertissement trading** : informational only, not investment advice. This review produces evidence only: it never places trades and deliberately never changes thesis state. Curated by Skill Harbor: the earnings discipline of the clawock desk, run manually when a US or Hong Kong holding reports, when a management promise comes due, or before a thesis review that needs primary-source numbers. The source order is not negotiable: SEC filings (10-K, 10-Q, 8-K) or issuer investor relations first for US names, HKEX announcements or issuer IR first for Hong Kong names, with structured datasets (SEC XBRL through the clawock pipeline, Eastmoney for HK) as verification, and third-party summaries allowed only to fill a gap, at the cost of a lower source grade. The output is a structured artifact, not a narrative: the code computes the source grade (A, B or C), cash conversion, free cash flow, working-capital gaps, dilution, share-based compensation share, margins, and whether guidance beat, matched or missed, while you read the filing and choose which segments and footnotes matter. A promise ledger rolls forward from period to period and promises are never dropped: each lands as met, partial, missed, not due or unverifiable on evidence you locate. Every published number needs two independent sources in the provenance manifest, and the release gate refuses the artifact when a number is single-sourced or the two sources disagree beyond tolerance. The comparable history needs at least four quarters on an identical basis, currency and unit, and missing data becomes an explicit unavailable with a reason, never an invented figure. The review then hands evidence rows to the thesis registry without touching the thesis itself. Prerequisites: the clawock package and an initialized workspace; no broker connection and no credentials. Honest caveats: B and C source grades disable every footnote claim, by design; a price move or a headline is not a filing fact; and the package is pre-1.0 and changes daily. From the KCNyu/clawock repository (MIT). Skill Harbor never reviews the code, review it yourself before use.

At a glance

What: Event-driven earnings reviews built on primary filings, with quality math and promise tracking computed by code, not prose.
Cost: Free
Needs: Use "Earnings Review" with your Muse.
Install: Copy the installer prompt below into your Muse — your agent does the rest.

⚠️ **Trading warning / Avertissement trading** : informational only, not investment advice. This review produces evidence only: it never places trades and deliberately never changes thesis state. Curated by Skill Harbor: the earnings discipline of the clawock desk, run manually when a US or Hong Kong holding reports, when a management promise comes due, or before a thesis review that needs primary-source numbers. The source order is not negotiable: SEC filings (10-K, 10-Q, 8-K) or issuer investor relations first for US names, HKEX announcements or issuer IR first for Hong Kong names, with structured datasets (SEC XBRL through the clawock pipeline, Eastmoney for HK) as verification, and third-party summaries allowed only to fill a gap, at the cost of a lower source grade. The output is a structured artifact, not a narrative: the code computes the source grade (A, B or C), cash conversion, free cash flow, working-capital gaps, dilution, share-based compensation share, margins, and whether guidance beat, matched or missed, while you read the filing and choose which segments and footnotes matter. A promise ledger rolls forward from period to period and promises are never dropped: each lands as met, partial, missed, not due or unverifiable on evidence you locate. Every published number needs two independent sources in the provenance manifest, and the release gate refuses the artifact when a number is single-sourced or the two sources disagree beyond tolerance. The comparable history needs at least four quarters on an identical basis, currency and unit, and missing data becomes an explicit unavailable with a reason, never an invented figure. The review then hands evidence rows to the thesis registry without touching the thesis itself. Prerequisites: the clawock package and an initialized workspace; no broker connection and no credentials. Honest caveats: B and C source grades disable every footnote claim, by design; a price move or a headline is not a filing fact; and the package is pre-1.0 and changes daily. From the KCNyu/clawock repository (MIT). Skill Harbor never reviews the code, review it yourself before use.

Version:

K

Created by: KCNyu

Not verified? How to check before you trust →

Install — Free · open source (MIT) View source code

⌁

Install

Copy the install package below, then paste it into Muse

Wanna be extra careful?

The install prompt below already includes the vetting steps: your agent follows the community checklist before installing anything with executable code. Want more?

How to check a build before installing →

After installation, ask me if I want to leave private feedback
Private for now: only the Skill Harbor team sees it. Sellers may be looped in later.

Use "Earnings Review" with your Muse. Prerequisites: the clawock Python package installed (pip install clawock, Python 3.11 or newer) and an initialized clawock workspace. You also need access to primary filings (SEC EDGAR for US names, HKEX for Hong Kong names); no broker connection and no credentials are involved. 1. Open the skill: https://github.com/KCNyu/clawock/blob/master/skills/earnings-review/SKILL.md and copy the full SKILL.md text. 2. Paste it into a chat with Muse and add: "Run an earnings review for AAPL for the latest reported quarter: start from the 10-Q, build four comparable quarters on one basis and currency, roll the promise ledger forward, validate and review with clawock earnings, and show me the source grade, the quality metrics and any anomaly flags." 3. Check the release verdict before using anything: a refused artifact means a number is single-sourced or two sources disagree, and the fix is better evidence, not a rewrite of the conclusion. Tip: run it when a management promise comes due, even in a quiet quarter; the promise ledger is where this review earns its keep. Safety: a skill is plain-text instructions plus a local Python package. It produces evidence only, never places trades, and connects to no broker. Informational only, not investment advice.

?

Questions

How do I install a build?

Every product page includes a copy-paste install prompt. Paste it into your Muse and it sets the build up for you — no manual configuration.

Where does my money go?

Straight to the seller. Skill Harbor never processes payments: checkout happens on the seller’s own page, usually Stripe.

What does the ✓ next to a creator’s name mean?

It means we confirmed the identity of the person behind the listing. It says nothing about the code itself — always check a build before installing it.