Release Change Brief

    1

    Compare two versioned structured snapshots with before/after values and source links. Keep unsupported differences and missing evidence visible instead of treating gaps as no change.

    Free

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Release Change Brief

    Example session with this skill installed

    Use Release Change Brief on the included wholly synthetic examples/complete.json. Compare demo-download-service 2.0 to 2.1 within the declared API and jobs scope, preserving source links and uncertainty. Return the complete Markdown brief. Do not fetch sources, publish release notes or approve a deployment.

    • Read your context and instructions
    • Compiled the release change brief

    Local verified synthetic example, not a live product release or Agensi execution. The unchanged standalone script produced the complete Markdown report below. Source eligibility and completeness are caller assertions, not independent verification.

    Release Change Brief

    demo-download-service: 2.0 → 2.1

    Evidence-linked snapshot changes found.

    Scope and coverage

    Coverage: complete_within_declared_scope

    • api: baseline=complete, eligible sources=baseline:api-inventory; target=complete, eligible sources=target:api-inventory
    • jobs: baseline=complete, eligible sources=baseline:jobs-inventory; target=complete, eligible sources=target:jobs-inventory
    • Captured: baseline=2026-01-10T12:00:00Z; target=2026-01-17T12:00:00Z

    Evidence-linked changes

    • item_added_to_snapshot: api / async-download
      • Before: {"present": false}
      • After: {"present": true}
      • Sources: baseline=baseline:api-inventory; target=target:api-inventory
    • attribute_changed: api / export / formats
      • Before: {"present": true, "value": ["csv"]}
      • After: {"present": true, "value": ["csv", "json"]}
      • Sources: baseline=baseline:api-inventory; target=target:api-inventory
    • attribute_changed: api / export / retry_limit
      • Before: {"present": true, "value": 2}
      • After: {"present": true, "value": 4}
      • Sources: baseline=baseline:api-inventory; target=target:api-inventory
    • item_removed_from_snapshot: api / old-download
      • Before: {"present": true}
      • After: {"present": false}
      • Sources: baseline=baseline:api-inventory; target=target:api-inventory
    • attribute_changed: jobs / dispatch / concurrency
      • Before: {"present": true, "value": 2}
      • After: {"present": true, "value": 4}
      • Sources: baseline=baseline:jobs-inventory; target=target:jobs-inventory

    Unverified differences

    • None established in this input.

    Gaps and follow-up

    • No gaps detected within the caller-declared scope.

    Source ledger

    • baseline:api-inventory | synthetic:2.0/api-inventory | version=2.0 | eligible
      • Supplied excerpt: Complete synthetic API inventory and complete fields: export has formats [csv], retry_limit 2, default_owner null. old-download has enabled true.
    • baseline:jobs-inventory | synthetic:2.0/jobs-inventory | version=2.0 | eligible
      • Supplied excerpt: Complete synthetic jobs inventory and fields: dispatch has concurrency 2.
    • target:api-inventory | synthetic:2.1/api-inventory | version=2.1 | eligible
      • Supplied excerpt: Complete synthetic API inventory and complete fields: export has formats [csv,json], retry_limit 4, default_owner null. async-download has enabled true and max_batch 100.
    • target:jobs-inventory | synthetic:2.1/jobs-inventory | version=2.1 | eligible
      • Supplied excerpt: Complete synthetic jobs inventory and fields: dispatch has concurrency 4.

    Limits

    • Only the supplied product, versions, domains, records and attributes are compared.
    • Evidence-linked means supplied, available, version-matched source excerpts; truth and relevance are not independently verified.
    • Completeness is a caller assertion, not proof that the underlying release is fully documented.
    • No deployment, compatibility, severity, security, customer impact or release-readiness conclusion is inferred.
    • Missing observations and missing evidence are not proof of no change.

    Input SHA-256: 8d82ab37d29614b12c4a6f8317648595ef275adfa03a329c11dbee5185b60b50

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

    About this skill

    What it does

    Prepare a reviewable release comparison from two structured snapshots of the same product. The skill matches stable domain and record IDs, preserves typed JSON values, and produces deterministic JSON or escaped Markdown with a source ledger and exact-input SHA-256.

    Evidence-linked changes, unverified differences and coverage gaps are reported separately. An omitted item or attribute becomes an addition/removal only when the opposite side declares complete coverage backed by eligible source excerpts. Missing sources, version mismatches, partial inventories and reversed capture times remain visible.

    This is an offline comparison of supplied assertions. It does not fetch release notes, crawl websites, inspect repositories, extract arbitrary documents, infer renames or independently verify source truth. A zero-change result does not mean the entire release is unchanged. No deployment, compatibility, severity, customer impact or release-readiness conclusion is made.

    Source and license

    Author: Evgeny Tkachenko. Copyright (c) 2026 Evgeny Tkachenko. MIT license included. Version 1.0.0. Fresh, company-agnostic implementation with wholly synthetic fixtures; no public GitHub repository or commit is claimed.

    Requirements

    Requirements: Python 3.10+, permission to read supplied inputs, write the unchanged standalone script to a temporary file, and execute it locally. No installation, dependencies, credentials or network connection. The script writes only stdout/stderr; the host may save requested report files.

    Verification

    Local verification covers the exact release ZIP, 83 tests and isolated synthetic examples. It does not establish suitability for production release decisions.

    On November 6, 2026, creator-account Agensi MCP delivery was verified. The delivered scripts matched the release ZIP and ran unchanged locally on macOS with Python 3.12.15. This was not Agensi-hosted execution, a non-creator buyer test or production validation. The delivery test checked complete, incomplete and stale synthetic inputs, Markdown output and malformed-JSON rejection. Incomplete and stale strict runs returned exit 3 with reports that retained their gaps.

    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.

    Only the supplied product, versions, domains, record IDs and attributes are compared. Does not fetch or independently verify sources, align renames, infer version order, perform semantic compatibility analysis or certify completeness. Source relevance/truth and coverage remain caller assertions. Missing evidence is not no change. Reports are not redacted. Input is limited to 2 MiB, nesting 24 and 50,000 values; invalid input fails instead of producing a comparison. Strict exit 3 still emits a report with gaps or unverified differences; exit 0 is processing success, not release approval. Verification covered creator-account MCP delivery and local execution only, not Agensi-hosted execution, non-creator access, production use or broader compatibility.

    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

    Fresh listing

    Recently published to Agensi

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 10 days ago

    • Free with an Agensi account

    Software
    Python 3.10+

    Permissions

    Terminal / Shell
    Read Files
    Write Files

    Reads caller-supplied sanitized structured snapshots and source excerpts, writes one temporary Python script, executes it offline, and optionally saves requested reports. Does not browse, contact services, read unrelated files, install dependencies, publish release notes, or deploy. Reports preserve supplied content; share only with separate authorization.

    Listed10 days ago

    What's inside

    Frequently Asked Questions

    Two explicitly versioned structured snapshots of the same product, using stable domain/record IDs and typed JSON attributes. It produces evidence-linked changes, unverified differences, coverage gaps and a source ledger, in JSON or Markdown.

    Provide the documented JSON request with product/version labels, capture times, coverage declarations, stable IDs, attributes and source excerpts. Python 3.10+ and local temporary-script/file access are required. No network, API key, installation or third-party package is needed.

    No. Completeness is conditional on supplied scope and assertions; evidence-linked does not mean independently verified. The skill does not infer compatibility, severity, customer impact or release readiness. Creator-account MCP delivery and unchanged local execution were verified on November 6, 2026 using synthetic inputs. This is not Agensi-hosted execution, an independent buyer test or production approval. MIT license, copyright 2026 Evgeny Tkachenko.

    New in Technical Documentation

    Free