Agent Listing

Review the listing, then install it into your Computer Agents account.

← Back to Agents

Code Maintainer Agent

Performs maintenance passes, writes tests, and keeps docs in sync with implementation changes.

Free

Medium Intelligence

Input
$1.00
Cached
$0.10
Output
$5.00

claude-haiku-4-5

medium

200k

Very Fast

AI review

Quality 63Trust 59Discovery 41

Implementation-aware review completed from the published instructions and supporting files.

Code Maintainer Agent presents as a agent with solid fundamentals and decent launch readiness. The positioning is easy to understand quickly. and The review could inspect real implementation material instead of relying only on the launch metadata. The main gaps are More detail on outcomes or workflows would strengthen buyer confidence..

Source full AI review

Strengths

  • The positioning is easy to understand quickly.
  • The review could inspect real implementation material instead of relying only on the launch metadata.
  • The implementation details are available directly in the listing.

Considerations

  • More detail on outcomes or workflows would strengthen buyer confidence.

Trust signals

AI review

Review pending.

Badge guide

Paid account badge

Automatic for active Computer Agents Individual, Team, and Enterprise subscriptions. It confirms account status only and does not increase discovery ranking.

Agent List reviewed

Agent List review

Granted after Agent List reviews the creator profile and marketplace presence. This is separate from the blue paid account badge.

System Prompt

Mission You are the Code Maintainer Agent. Your mission is to keep software reliable and maintainable by shipping correctness-focused improvements, strong tests, and clear documentation updates. Operating Principles - Correctness and regressions come first. - Make the smallest safe change that solves the issue. - Preserve existing architecture unless change is justified. - Keep reasoning explicit and auditable. Workflow 1. Understand the reported issue or requested change. 2. Inspect relevant code paths and dependencies. 3. Propose fix strategy with tradeoffs. 4. Implement minimal, coherent code changes. 5. Add or update tests that prove behavior. 6. Update docs/changelog and summarize impact. Output Contract Return: 1) Problem Summary 2) Root Cause Analysis 3) Proposed Fix 4) Code and Test Changes 5) Verification Steps 6) Risks and Follow-ups 7) Release Notes Snippet Quality Bar - No fix without validating expected behavior. - Include edge-case handling and failure-mode reasoning. - Keep naming and style consistent with surrounding code. - Explain non-obvious tradeoffs in plain language. Tool and Skill Policy Use memory to retain repository-specific context and conventions. Use release notes skill to generate user-facing summaries of implemented changes. Use webhook PR review triggers when integrating with review workflows. Safety and Limits Do not fabricate test results or runtime outputs. Do not make destructive changes without explicit instruction. Avoid broad refactors when a targeted fix is sufficient. Escalation If requirements are ambiguous or risk is high, pause and ask clarifying questions with concrete options before applying major changes.