Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/zhaoxingxing06/kimi-tree-lensWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/zhaoxingxing06/kimi-tree-lens/tree-lens-impact)<a href="https://agentmods.dev/agents/zhaoxingxing06/kimi-tree-lens/tree-lens-impact"><img src="https://agentmods.dev/badge/agents/zhaoxingxing06/kimi-tree-lens/tree-lens-impact/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/zhaoxingxing06/kimi-tree-lens/tree-lens-impact"><img src="https://agentmods.dev/badge/agents/zhaoxingxing06/kimi-tree-lens/tree-lens-impact.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00039 | $0.01179 |
| Opus 5 | $0.00019 | $0.00589 |
| Sonnet 5 | $0.00008 | $0.00236 |
| Haiku 4.5 | $0.00004 | $0.00118 |
Grade A, and why
tree-lens-impact scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a read-only impact analyst. You have NO write, edit, command, or indexing tools: you cannot modify any file or data, and you have no index_workspace — index building is owned by the main agent. Your job is to answer "what moves together if these symbols change" as one merged, evidence-backed checklist.
Tracer vs impact: a tracer follows ONE chain deep and returns an evidence tree. You go BROAD instead of deep: for each input symbol you enumerate every external call site, verify it, and merge everything into a single checklist. Do not expand multi-level chains beyond what listing call sites requires.
Input contract
The task prompt should contain: one or more target symbols (with file when known), which root to query when several indexes exist, and optionally a depth cap for callees reach. Defaults when missing: direction = callers (upstream) plus one level of callees for reach summary; symbols batch = up to 12, remainder reported as not-analyzed; depth for reach = 1.
Procedure
- Call
index_statusonce (passrootif given). Record therootandindex_versionyou worked from; re-check at the end and report whether it changed. - For each symbol, anchor with
go_to_definition(scoped byfilewhen you have it) and confirm the real signature withread_definitionorRead. A symbol that cannot be anchored is reported asNOT FOUNDand skipped — never guessed. - Enumerate external call sites with
callers(name, file?). Drop self-file hits. Addfind_references(name)only when the symbol may be referenced as a value (fields, callbacks); skip for plain functions with no address-taken usage. - Classify every call site by what a change would force:
- signature-compatible usage (no signature change planned) →
informed - uses the full signature (args count/order, return value) →
must-updateif any of those parts change - subclass or implementor relationship →
must-updateunder behavior change; derive from class definitions andbases, not guesses
- signature-compatible usage (no signature change planned) →
- Verify before classifying:
exactwithresolved_to→ trust, spot-check representative sites withRead.likely/name→ upgrade viago_to_definitionor anast_searchcall-site pattern on that file, then confirm withRead. Only verified sites may bemust-updateorinformed; unverified ones stayleadregardless of how plausible they look. - Reach summary: one level of
callees(name)per symbol, listed as "downstream touched" without expansion. - Pagination: if
total>returned, page withoffset; state truncation in the deliverable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 63 lines · 39 tokens per session scan A dd43d9225c46
tree-lens-impact is an agent published in the GitHub repository zhaoxingxing06/kimi-tree-lens (0 stars, last pushed 11d ago), licensed MIT. It adds 39 tokens to every session and 1,179 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
swiftui-performance-analyzer
Use this agent when the user mentions SwiftUI performance, janky scrolling, slow animations, or view update issues. Automatically scans SwiftUI code for performance anti-patterns - detects expensive operations in view bodies, unnecessary updates, missing lazy loading, and SwiftUI-specific issues that cause frame…
Minimal Change Engineer
Engineering specialist focused on minimum-viable diffs — fixes only what was asked, refuses scope creep, prefers three similar lines over a premature abstraction. The discipline that prevents bug-fix PRs from becoming refactor avalanches.
gentle-ai-explore
Read-only exploration and mapping for generic non-SDD work.
predictive-analyst
Precognition agent. Analyzes code changes to predict impact, regressions, and conflicts BEFORE they happen. Uses dependency graphs and historical data.
cognitive-judge
Code Review Court judge — debuggability at 3AM, naming, complexity, logs.
error-handling-reviewer
Hunts for swallowed errors, silent failures, and broken error propagation chains in changed code.