Shihwesley Mercator Ai
shihwesley-mercator-ai is a Claude Code plugin for the Build phase that provides Merkle-enhanced codebase mapping with O(1) change detection and zero-token staleness prevention for AI agents.
by shihwesley · github.com/shihwesley/mercator-ai
Install Mercator so Claude Code agents get Merkle-backed codebase maps, manifest files, and zero-token staleness guards when the repository changes.
Add it to Claude Code
Install the plugin in Claude Code. One command, paste-ready.
/plugin install shihwesley-mercator-ai@shihwesley/mercator-aiBuilt to be called by your agent
Skillselion is itself an MCP server. Your agent can pull this entry and a paste-ready install config straight from the API - no copy-paste.
Retrieve this entry with skillselion.get_details("plugin:shihwesley/mercator-ai") and the paste-ready config with skillselion.get_install_config("plugin:shihwesley/mercator-ai").
What it does
shihwesley-mercator-ai is a Claude Code plugin that builds and maintains a Merkle-enhanced map of your codebase so AI agents spend tokens on what actually changed. Solo builders juggling Cursor, Claude Code, and subagents routinely hit stale context: the agent confidently edits the wrong module because its mental model of the repo is outdated. Mercator targets that failure mode with change detection described as O(1) relative to naive full-tree reanalysis, plus explicit staleness prevention so documentation manifests do not silently drift. The workflow emphasizes codebase mapping, architecture analysis, and onboarding—keywords point at outputs like a codebase map markdown file and a mercator JSON manifest suitable for agent consumption. Use it when your repo outgrows casual @-folder prompts, when you spawn parallel subagents, or when you iterate frequently in Operate and need maps to stay current without paying full re-read costs every session.
Highlights
- Merkle-enhanced codebase mapping with O(1) change detection positioning
- Zero-token staleness prevention so agents do not re-ingest unchanged regions
- Produces structured artifacts such as docs/codebasemap.md and docs/mercator.json per plugin keywords
- Supports parallel subagent onboarding to large repositories
- Single-plugin developer-tools bundle focused on mapping and architecture analysis—not generic chat
Why builders use it
Coding agents waste context and make wrong edits because full-repo rescans are expensive and codebase documentation goes stale after every refactor.
After install, agents can rely on updated mercator manifests and codebase maps that refresh incrementally when files change, cutting redundant token use and wrong-file navigation.
At a glance
- Type - Plugin in Development Tools.
- Adoption - 0 installs, 4 stars, 0 votes.
FAQ
Who is shihwesley-mercator-ai for?
It is for developers who agent-code in Claude Code and need structured repo maps and manifests that stay current as the codebase evolves.
When should I use shihwesley-mercator-ai?
Use it at the start of a Build sprint to onboard agents to the repo, and again after significant refactors when staleness would mislead subagents.
How do I add shihwesley-mercator-ai to my agent?
Install the shihwesley/mercator-ai Claude Code plugin, run its mapping workflow in your project root, and point agents at the generated codebase map and mercator JSON artifacts.
Comments
Share how you use shihwesley-mercator-ai, gotchas, or tips for other indie builders.
No comments yet - be the first to share how you use it.