- Home
- Skills
- Code Quality & Review
- Change Impact Map
Change Impact Map
Trace potential change impacts through a supplied dependency graph, with source-linked paths and explicit coverage gaps.
Free
Works with the AI Agents you already use
Change Impact Map
Example session with this skill installed
Use Change Impact Map on the supplied fictional JSON graph. Trace the order-schema change, explain the evidence behind each potential impact, report cycles and limits, and suggest follow-up checks without claiming they were executed.
- Read your context and instructions
- Compiled the change impact map
Local verified synthetic graph example. The unchanged offline script produced this complete output from an invented dependency map. No real system was inspected, no source locator was fetched, and no suggested follow-up test was executed.
Change Impact Map
Potential dependency exposure, not a release approval.
Changed: order-schema
Coverage: partial; Fictional order-processing example only
Traversal truncated: false
Source date or revision: synthetic-demo-revision-17
Potential impacts
- Daily export (daily-export): unsupported
- Path: order-schema → daily-export
- Edge IDs: schema-export
- Evidence IDs: none
- Orders API (orders-api): supported
- Path: order-schema → orders-api
- Edge IDs: schema-api
- Evidence IDs: src-api
- Web checkout (web-checkout): supported
- Path: order-schema → orders-api → web-checkout
- Edge IDs: schema-api, api-web
- Evidence IDs: src-api, src-web
Cyclic groups
- Nodes: orders-api, web-checkout; edge IDs: api-web, web-api-callback
Edge evidence
- api-web: web-checkout depends on orders-api; observed; evidence: src-web
- schema-api: orders-api depends on order-schema; observed; evidence: src-api
- schema-export: daily-export depends on order-schema; inferred; evidence: none
- web-api-callback: orders-api depends on web-checkout; observed; evidence: src-callback
Source locators (not fetched or verified)
- src-api: fixture://demo-rev-17/orders-api/schema-import; Synthetic API imports the order schema.
- src-callback: fixture://demo-rev-17/orders-api/checkout-callback; Synthetic API models a dependency on a checkout callback.
- src-web: fixture://demo-rev-17/web-checkout/api-client; Synthetic checkout calls the orders API.
Limits and unknowns
- Potential impact only: reachability does not prove a behavior change or failure.
- Completeness and evidence truth are not independently verified; absence of a path is not proof of no impact.
- One deterministic shortest discovered witness per node; alternate paths and changed roots are not enumerated.
- Cycles are reported only within visited nodes and inspected edges; uninspected cycles can be absent.
- The supplied dependency map is incomplete or its coverage is unknown.
- Some witness paths contain inferred, unknown, or uncited edges; verify them before relying on those paths.
- Declared gap: External consumers and runtime feature flags are not modeled.
- Declared gap: The daily-export relationship is an uncited assumption.
- max_depth: 6
- max_nodes: 200
- max_edges: 2000
- depth_limited_nodes: []
- node_budget_omitted_nodes: []
- edge_budget_reached: false
- nodes_with_uninspected_edges: []
Connects securely to your tools. The creator never sees your data.
About this skill
What it does
Turn changed component IDs and a small JSON dependency graph into a bounded impact review. Identify directly and transitively exposed components, retain one shortest discovered evidence path per node, distinguish supported, inferred, and unsupported paths, and report cycles found in inspected relationships. JSON and Markdown reports preserve graph scope, declared gaps, unknown changed IDs, and traversal limits. Useful for change review, compatibility questions, and test planning.
This is advisory analysis of caller-supplied graph data. A supported path means its supplied edges have evidence labels, not that real-world dependency truth or impact has been independently verified. Not reached does not mean unaffected. Missing coverage, uncited relationships and traversal limits stay visible.
Source and license
Author: Evgeny Tkachenko. Copyright (c) 2026 Evgeny Tkachenko. MIT license included and existing contributor notices retained. Package version 1.0.1 is an attribution-only revision; runtime version 1.0.0 and behavior remain unchanged. No public GitHub repository or commit is claimed.
Requirements
Requires Python 3.9+ standard library and permission to read supplied graph input, write an unchanged temporary script and execute it locally. No network, credentials, installation or third-party dependency is needed.
Verification
Local verification covers 23 bundled tests, including 40 seeded synthetic graph cases, the exact archive and full/capped example runs. This is not a real-system dependency audit or release approval.
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 the supplied full and depth-capped graphs, Markdown output and rejection of a negative depth. The capped run retained its truncation and frontier warnings.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Code Quality & Review.
Advisory analysis of supplied graphs only; no automatic repository dependency discovery.
Potential exposure is not proof of a behavior change, failure, severity, or release safety.
Sources, timestamps, relationships, and real-world graph completeness are not independently verified.
Traversal is capped; cycles are limited to visited nodes and inspected edges.
Only one shortest discovered witness is retained per reached node, not all paths or changed roots.
Python versions other than 3.12 were not executed in this local verification.
A supported path refers to supplied labels/evidence, not an independent source audit. Missing graph coverage, uncited assumptions and traversal truncation remain uncertainty. 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
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
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 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 3.9+
Permissions
Reads only supplied graph data, writes and executes one unchanged temporary Python script, and optionally saves requested outputs. Does not discover repository dependencies, follow source URLs, use credentials, contact services, execute suggested checks, or change code/deployment state.