Is the Understand Chat skill free?

Yes. The skill itself is free from Egonex-AI/Understand-Anything. SkillProof publishes the install command and an independent test verdict at no cost.

What is the Understand Chat SkillProof Score?

8.0/10 — installs cleanly 5/5, triggers reliably 4/5, output vs. baseline 7/10, docs & honesty 4/5.

How do I install Understand Chat?

Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.

Can I use Understand Chat with Cursor, Copilot, Gemini CLI, Codex or other AI tools?

The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.

Understand Chat

Answers codebase questions by grepping a pre-built knowledge-graph.json instead of re-reading the whole repo.

by Egonex-AI · Egonex-AI/Understand-Anything

Works with setup

Install — copy the command Download source ↗

Test report

Verdict: Works with setup
Score: 8.0/10
Tested: Jul 10, 2026
Environment: Claude Code 2.x (agent harness)
Upstream re-checked: Aug 10, 2026 · e12e917

Re-tested against upstream HEAD (2026-07-17): the authors added a graph-freshness gate that resolves the graph's gitCommitHash, diffs it against HEAD with a project-scoped pathspec, and warns before answering from stale data. I executed that bash block verbatim on a fixture repo whose graph predated an MFA commit and an uncommitted edit -- it correctly flagged all three drifted files, so the skill no longer answers from a stale graph silently. The graph-structure reference in SKILL.md now matches the field names in both the producer pipeline and packages/core types, resolving the doc mismatch we originally flagged. What has not changed: it queries a knowledge graph only the sibling /understand skill can generate, so installed alone it still has nothing to work with. Stage 0 clean -- the new bash is read-only git plumbing with no network calls.

Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works

Installs cleanly 5/5
Triggers reliably 4/5
Output vs. baseline 7/10
Docs & honesty 4/5

What Understand Chat does

A companion skill to /understand: it greps the project's knowledge-graph.json (nodes, edges, layers) for query keywords, follows the 1-hop edge subgraph, and answers using only that slice instead of dumping the full graph into context. Triggers on codebase-understanding questions ('what depends on X', 'explain this architecture using the graph') but requires the sibling /understand skill to have already generated the graph file.

How to install Understand Chat

git clone https://github.com/Egonex-AI/Understand-Anything
cd Understand-Anything
mkdir -p ~/.claude/skills
cp -r understand-anything-plugin/skills/understand-chat ~/.claude/skills/understand-chat

Skills live in ~/.claude/skills/ (global) or .claude/skills/ (per-project). Restart Claude Code after installing.

Commands — how to trigger Understand Chat

/understand-chat Answers codebase questions by grepping a pre-built knowledge-graph.json instead of re-reading the whole repo.

It also activates on plain-language prompts like these:

What depends on this module, according to the knowledge graph?
Explain this architecture using the project's knowledge-graph.json
Grep the codebase graph for callers of this function instead of reading everything

Frequently asked questions

Is the Understand Chat skill free?: Yes. The skill itself is free from Egonex-AI/Understand-Anything. SkillProof publishes the install command and an independent test verdict at no cost.
Does Understand Chat work with Claude Code?: We tested it with Claude Code 2.x (agent harness) on Jul 10, 2026. Verdict: Works with setup. Re-tested against upstream HEAD (2026-07-17): the authors added a graph-freshness gate that resolves the graph's gitCommitHash, diffs it against HEAD with a project-scoped pathspec, and warns before answering from stale data. I executed that bash block verbatim on a fixture repo whose graph predated an MFA commit and an uncommitted edit -- it correctly flagged all three drifted files, so the skill no longer answers from a stale graph silently. The graph-structure reference in SKILL.md now matches the field names in both the producer pipeline and packages/core types, resolving the doc mismatch we originally flagged. What has not changed: it queries a knowledge graph only the sibling /understand skill can generate, so installed alone it still has nothing to work with. Stage 0 clean -- the new bash is read-only git plumbing with no network calls.
What is the Understand Chat SkillProof Score?: 8.0/10 — installs cleanly 5/5, triggers reliably 4/5, output vs. baseline 7/10, docs & honesty 4/5.
How do I install Understand Chat?: Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
Can I use Understand Chat with Cursor, Copilot, Gemini CLI, Codex or other AI tools?: The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.