Works with the AI Agents you already use
Uncovr
Reconstruct features, business rules and change risks from a legacy repository, with source evidence and human review checkpoints.
$5
/moSubscription · cancel anytime
Secure checkout via Stripe
- Always the latest version
- Delivered live to your agent via MCP
- Cancel anytime, access ends at period end
Uncovr
Example session with this skill installed
Analyze the /src/billing module in this repo. I need to understand the discount logic and subscription states before we migrate to the new API. Map it to Obsidian.
- Read your context and instructions
- Compiled the uncovr
I have initialized the product context for the /src/billing module. I analyzed 14 files and identified 8 core business rules regarding discount stacking and subscription transitions.
Billing Module Baseline: Initial Scope
| Artifact | Status | Findings |
|---|---|---|
overview.md | Created | Scope: Subscription logic, Discount engine. |
features.md | 3 Identified | Tiered pricing, Coupon validation, Grace period handling. |
domain.md | 5 Entities | Subscription, Plan, Transaction, Discount, UserRole. |
evidence.jsonl | 12 Claims | Traceable links to src/billing/engine.py and src/billing/models.py. |
Reconstructed Business Rules (Draft)
- BR-001: Discounts are non-stackable by default (see
engine.py:L42). - BR-002: Subscriptions entering
PAST_DUEstate trigger a 3-day grace period before suspension.
BR-003: Pro-rated refunds are calculated based on remaining days in the billing cycle, excluding the day of cancellation.
Next steps
- Review the proposed scope in
product-context/overview.md. - Confirm the
wiki_target: obsidiansetting to enable OKF export. - Approve the initial claim set to proceed to Stage 2 (Architecture Mapping).
Subscription · always the latest version
- Secure checkout via Stripe
- Cancel anytime
- Delivered live via MCP
About this skill
Inheriting a legacy codebase usually means guessing at business logic hidden in undocumented files. This skill reconstructs a verifiable product baseline from your existing code, mapping features and business rules with evidence before you touch a single line.
What it does
- Feature Reconstruction identifies core capabilities and user journeys by analyzing code patterns and history.
- Business Rule Recovery extracts implicit logic and constraints into a machine-readable evidence index.
- Hotspot Mapping identifies modules with high change frequency or bug density using Git history analysis.
- Traceability Linking connects every reconstructed claim to specific lines of code or commits via JSONL evidence.
- Baseline Generation produces a versioned set of six standard Markdown artifacts and an optional LLM-ready wiki.
How it works
- Initialize and Scope: The skill scans the repo, proposes an analysis batch, and waits for your confirmation on scope and wiki targets.
- Explore and Map: It builds an inventory of the system architecture and module dependencies within the agreed boundaries.
- Analyze and Document: Reconstructs the domain model and features, recording stable claim IDs and source evidence.
- Validate and Review: Runs an optional independent subagent review to flag gaps or contradictions before a human checkpoint.
Frameworks & tools
Requires a terminal with python3 3.10+ and git. It utilizes custom Python scripts for structural checks, history analysis, and Open Knowledge Format (OKF) wiki exports.
Why this beats prompting it yourself
Generic prompts often hallucinate features or miss subtle business rules buried in edge cases. This skill enforces a strict evidence contract, requiring every claim to be backed by a source pointer, and includes built-in structural validation scripts to ensure the documentation is technically sound.
Use cases
- Legacy Takeover when you need to understand a system with zero documentation.
- Refactoring Preparation to map business rules before a major architectural change.
- Product Baselining to recover context for product owners after developer turnover.
Known limitations
Does not support code modification, test writing, or pull request reviews. Requires manual human approval for business validation and final baseline status.
Requires Python 3.10+ and Git. Read-only analysis; cannot modify source code or run destructive commands. Cannot approve claims without human intervention.
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 any payment.
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 forever
No account required to browse
Trust & safety
Security scan
Security scan passed 4 days ago
Requires