- Home
- Skills
- Agents & Orchestration
- Agent Workflow Skills Pack
Agent Workflow Skills Pack
Automate code reviews, debugging, dependency upgrades, and documentation sync with a verified workflow toolkit.
$12
Works with the AI Agents you already use
Agent Workflow Skills Pack
Example session with this skill installed
Review my recent changes for bugs and security issues, then write a PR description for them.
- Read your context and instructions
- Compiled the agent workflow
I have reviewed the changes in the current branch against main and prepared a PR description.
Code Review Findings
| Severity | Finding | Location | Failure Scenario | Suggested Fix |
|---|---|---|---|---|
| Critical | SQL Injection | db/query.js:42 | User input from req.query.id is passed directly to template string. | Use parameterized queries with db.execute(). |
| High | Unhandled Promise Rejection | api/user.js:15 | fetch() call lacks a .catch() or try/catch block. | Wrap in try/catch and return a 500 status. |
| Medium | Missing Auth Check | routes/admin.js:8 | Exported route lacks the requireAdmin middleware present in others. | Add auth.requireAdmin to the middleware chain. |
Verdict: needs rework
PR Description
Why
Fixes a critical SQL injection vulnerability in the user lookup flow and addresses unhandled errors in the API layer.
What changed
- Security: Switched to parameterized queries in
db/query.js. - Error Handling: Added try/catch blocks to user API endpoints.
- Middleware: Added missing admin authorization check to admin routes.
How it was tested
- Ran
npm test(all 12 tests passed). - Manually verified SQL injection payload
' OR 1=1 --is now escaped.
Risks & rollout
- Low risk. The change is behavior-preserving for valid inputs.
Next steps
- Apply the suggested fix for
db/query.js. - Run the new regression test suite for the auth check.
- Use
gh pr createto submit these changes.
Connects securely to your tools. The creator never sees your data.
About this skill
Maintaining high-quality codebases requires constant vigilance across debugging, testing, and documentation. This Agent Workflow Skills Pack provides a specialized toolkit to automate the most repetitive and error-prone parts of the development lifecycle, from deep-dive debugging to safe dependency management.
What it does
- Code Audit reviews diffs for correctness, security vulnerabilities, and maintainability issues with line-specific evidence.
- Root Cause Analysis reproduces bugs, forms ranked hypotheses, and implements verified fixes for failing tests or crashes.
- Dependency Management inventories outdated packages and performs incremental, batch-tested upgrades to prevent breaking changes.
- Documentation Sync identifies stale READMEs, docstrings, and API references by analyzing code changes and updating them to match reality.
- Refactoring executes behavior-preserving restructures through small, verified steps with continuous test validation.
- Automated Reporting generates structured PR descriptions, release notes, and regression tests based on actual git history.
How it works
- Analyze: The agent scans the current git state, diffs, or lockfiles to understand the technical context.
- Execute: It runs specialized workflows like
debug-investigationortest-writerbased on your specific command. - Verify: Every change is validated against the test suite, type checker, or build system to ensure zero regressions.
- Report: Findings and changes are delivered in severity-ranked lists, Markdown tables, or standard changelog formats.
Frameworks & tools
Works with standard package managers (npm, yarn, pnpm, pip, uv, poetry, cargo, go mod, bundler) and git workflows. Supports any language with a test suite and is compatible with GitHub CLI (gh).
Why this beats prompting it yourself
Generic prompts often suggest "guess-fixes" that hide symptoms rather than solving root causes. This skill enforces a strict verification loop, requiring reproduction of bugs and green tests before any change is finalized.
Use cases
- Auditing a complex PR for security sinks and off-by-one errors before merging.
- Upgrading a project with dozens of outdated dependencies without breaking the build.
- Updating documentation and example code after a major API refactor.
- Writing regression tests for a bug fix that match the project's existing mocking style.
Known limitations
Does not perform style-only linting. Requires a working local test environment to verify fixes and refactors. Package upgrades require network access to registries.
Requires a working test suite and local environment for verification. Does not handle pure style/linting fixes. Dependency upgrades require registry access.
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
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 3 days ago
Requires