Aegis Markdown Blueprint & Integrity Master

    1

    It turns a user's brief and approved local source files into a reader-focused blueprint, draft, and review record.

    Free

    2 usesSecurity scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Aegis Markdown Blueprint & Integrity Master

    Example session with this skill installed

    Use only these fictional source facts: the current charge endpoint is POST /v2/charges; authentication uses bearer tokens that expire after 24 hours; each API key is limited to 600 requests per minute; failed webhooks retry 5 times with exponential backoff; and /v1/charge is deprecated with no removal date decided. Create a concise blueprint and a complete Markdown integration guide for external developers. Keep unknowns visible. Do not invent a base URL, payload, test results, or contact address.

    • Read your context and instructions
    • Compiled the aegis markdown blueprint

    Blueprint snapshot

    • Reader: External developers integrating the payments API for the first time.
    • Outcome: Make a charge request and understand the documented limits and migration status.

    Source map: Authentication and token lifetime; current charge endpoint; rate limit and webhook retry behavior; deprecated endpoint and unknown removal date.

    • Section plan: Quick start; Limits and retries; Migration.
    • Open items: The /v1/charge removal date and a public integration contact address were not supplied.

    payments-api-integration-guide.md

    Payments API integration guide

    Quick start

    Use a bearer token to authenticate. Tokens expire after 24 hours. Create a charge with POST /v2/charges. Replace <BASE_URL> only with the gateway URL approved for your environment; no base URL or request payload was supplied.

    Limits and retries

    Each API key is limited to 600 requests per minute. Failed webhooks are retried 5 times with exponential backoff.

    Migration

    The /v1/charge endpoint is deprecated. Its removal date has not been decided, so confirm the date before setting a migration deadline.

    Open items

    • Confirm the approved gateway URL and request payload from your API documentation.
    • Confirm the removal date for /v1/charge.
    • Confirm the public contact address for integration questions.

    Included example check

    The package's synthetic example check report records PASS with 0 errors and 0 warnings for its structural and selected-literal checks. This result does not certify factual correctness or semantic completeness.

    Next steps

    Confirm the three open items from authoritative sources, then rerun the checker and review the guide before sharing it.

    payments-api-integration-guide.md

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Build Markdown around what its reader needs

    Blueprint MD is a standalone, English-first skill for people creating Markdown for their own readers. It turns a user's brief and approved local source files into a reader-focused blueprint, draft, and review record. It does not require Aegis Stable, private Aegis code, or a creator-specific workflow. The buyer-facing workflow is English-first; source material may be in other languages.

    What it does

    • Clarifies the reader, outcome, scope, source files, and acceptance checks before drafting.
    • Maps must-have requirements to planned sections and supporting sources.
    • Drafts the requested Markdown while keeping unanswered questions visible.
    • Runs local checks for missing or unmapped requirements, missing or empty sections, selected document-type structure, visible open questions, and selected literal details that are absent from supplied sources or an explicit allowlist.
    • Supports a review pass after edits so the agent can compare changed sections with the blueprint.
    • Includes patterns for README files, guides, runbooks, and specifications.

    Optional conversion and cleanup helpers

    When requested, the included helpers can support a task such as converting docs/api.md to HTML with a compatible converter already installed, normalizing heading levels into a new copy, and reporting missing local link targets or same-file anchors. The inventory identifies external URLs but does not fetch or validate them. Conversion depends on a suitable local converter; this skill does not download or install one. These are supporting workflows. Reader-first Markdown planning, drafting, and review remain the core.

    Included

    An English-first blueprint template and playbook; a Python standard-library checker; synthetic before-and-after examples; and helpers for Markdown inventory, copy-only normalization, converter detection, limited conversion checks, and ZIP packaging with hashes.

    Typical outputs

    The requested Markdown document, a concise blueprint, a checker report when the checker is run, and an open-items file when questions remain. The agent should report exactly which files it created and which checks it ran.

    Requirements

    A compatible SKILL.md agent with user-authorized local file and terminal access. Python 3.9+ standard library for bundled checks. Optional format conversion requires a compatible converter already installed. The bundled tools do not require network access or install software.

    Limits

    The checker covers selected structures and literal patterns; it can miss unsupported claims and flag harmless details. A matching source token does not prove correctness, and a heading map does not prove prose satisfies a requirement. The checks do not certify truth, reasoning, semantic consistency, readability, legal sufficiency, security, or whether a procedure works. Runbook labels are not executed. Local link targets and selected same-file anchors can be checked; external links are not fetched or validated. Conversion checks are limited and heuristic, and some formats need visual review. Human review remains necessary.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Technical Documentation.

    The checker covers selected document structures and literal patterns only. It may miss unsupported claims and flag harmless details. Matching text in a source does not establish truth or correct use. Structural checks do not prove semantic completeness, readability, security, legal sufficiency, or operational correctness. Runbook labels are not executed. The inventory can report missing local link targets and same-file anchors; external URLs are not fetched or validated. Optional conversion depends on a compatible converter already installed and may need visual review. Human review is required.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    2 uses

    Times used to date

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 13 days ago

    • Free with an Agensi account

    Software
    Terminal accessPython 3.9+
    Optional
    Format converter

    Python runs the bundled checks. Optional format conversion needs a compatible converter already installed.

    Permissions

    Read Files
    Write Files
    Terminal / Shell

    Use only user-approved source files and the output folder the user chooses. Source files remain unchanged by default; new drafts, blueprints, and reports are written only when requested. No browser, network, external accounts, environment variables, or automatic software installation are required.

    Listed16 days ago
    Updated13 days ago

    What's inside

    Frequently Asked Questions

    It can produce the requested Markdown document, a concise blueprint mapping must-have points to planned sections and sources, and a check report when the bundled checker is run. An open-items file is useful when questions remain. The agent should report which files it actually created.

    It checks blueprint fields and IDs, requirement-to-heading mappings, missing or empty planned sections, extra headings, selected README/runbook/spec structure, visible open questions, and selected URL, email, date, version, endpoint, path, port, status-code, and quantity literals against supplied sources or an exact allowlist.

    No. It checks only selected structures and literal patterns. It can miss unsupported statements and flag harmless details. A source match does not prove a statement is true or used correctly. Human review is still needed.

    The bundled checks use Python 3.9+ standard library and need no network access. Optional format conversion requires a compatible local converter that is already installed. The skill does not download or install software.

    No by default. It reads only the source scope the user approves and writes new deliverables to the output location the user chooses. The normalization helper writes to a separate output file and refuses in-place changes.

    It includes optional local helpers. With a compatible converter already installed, an agent can convert Markdown to a supported format, normalize selected properties into a separate copy, and inventory missing local link targets and same-file anchors. External URLs are not fetched or validated. These helpers support the core reader-first blueprint and review workflow; the skill does not install converters.

    New in Technical Documentation

    Free