knowledge-capture_skill

This skill captures and organizes business rules, patterns, and interfaces discovered during analysis, ensuring deduplication and proper categorization.
  • Shell

168

GitHub Stars

2

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill rsmdt/the-startup --skill knowledge-capture

  • reference.md11.1 KB
  • SKILL.md7.8 KB

Overview

This skill captures and organizes business rules, technical patterns, and external interface contracts discovered during analysis or implementation. It enforces a three-folder documentation hierarchy (domain, patterns, interfaces) and mandates deduplication before creating new files. The goal is searchable, maintainable docs that get updated rather than duplicated.

How this skill works

When you discover reusable logic, integrations, or implementation patterns, run the deduplication checks against docs/ first. If existing content covers the topic, update that file; otherwise determine the correct category (domain, patterns, interfaces), create a descriptive filename, apply the appropriate template, and add cross-references. The skill also records a short report of files created or updated.

When to use it

  • When you find business rules, workflows, or domain validation logic
  • When you discover reusable technical or architectural patterns
  • When you integrate with third-party APIs, webhooks, or external services
  • When you add or change public API contracts or service interfaces
  • Before opening a PR that introduces reusable knowledge or breaking changes

Best practices

  • Always perform deduplication searches (grep/find) before creating a new doc
  • Prefer updating an existing, related document instead of creating a duplicate
  • Use clear, searchable filenames (avoid vague names like api.md or auth.md)
  • Follow the template for the chosen category: domain, pattern, or interface
  • Include title, context, details, examples, references, and cross-links in every doc
  • Add the required end-of-task report listing docs created/updated and the action taken

Example use cases

  • Implementing Stripe or SendGrid integration → create docs/interfaces/stripe-payments.md and use interface template
  • Discovering a common caching approach in multiple modules → update or create docs/patterns/caching-strategy.md
  • Defining user permission rules discovered during analysis → update docs/domain/user-permissions.md
  • Noticing a new public endpoint or changed signature → create/update interface doc and flag for cross-reference updates
  • Finding a new validation rule for orders → add to docs/domain/order-validation.md and link related workflow docs

FAQ

If related docs exist and your new info enhances or narrows the topic, update the existing file. Create a new file only when the topic is distinct enough to avoid confusion.

How do I name a document?

Pick a descriptive, searchable filename that reflects intent and scope, e.g., stripe-payment-integration.md or authentication-flow.md; avoid abbreviations or vague names.

What tools should I use to deduplicate?

Use grep -ri and find across docs/ to search by keywords or filename patterns. Read candidate files fully to confirm overlap before deciding.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational