How do I connect TraceWeave MCP Server to Claude?

The listing records ` /.venv/bin/python /server.py` as its setup step. Run it, then follow the repository's instructions for the client configuration; the listing names Claude Code, Cursor as compatible clients.

Is TraceWeave MCP Server free?

The listed licence is MIT. Check the upstream terms for permitted use and commercial requirements; a public repository does not by itself mean the software is free or open source. Connected APIs and hosted services may have separate charges.

What can TraceWeave MCP Server do?

TraceWeave MCP Server documents 22 tools to the agent, including analyze_failure_event, analyze_failures, build_kdb. The descriptions above come from project documentation. A live handshake does not test individual tool behavior.

TraceWeave

Unclaimed

by gokeshenzhen

Open-source, evidence-driven MCP server for RTL simulation debugging: correlate VCS/Xcelium logs, VCD/FSDB waveforms, SystemVerilog/UVM source, hierarchy, and connectivity to trace failures to root cause.

Install

$<TRACEWEAVE_HOME>/.venv/bin/python <TRACEWEAVE_HOME>/server.py

Set up this server

Compare servers My Stack

This server needs project-specific setup. Follow the project instructions; no reusable public launch command is available yet.

Project instructions

chip-verificationedafsdbhardware-verificationllm-toolsmcpmcp-servermodel-context-protocolpyslangroot-cause-analysisrtlrtl-verificationsimulation-debuggingstatic-analysissystemveriloguvmvcdverdiverilogwaveform

Is this your MCP server?

This listing was auto-indexed from the public record. Claim it to edit the page, set compatibility and unlock growth tools. Takes under two minutes.

Claim this server

Security profile

Claimed and verified servers get a weekly static scan that shows what the code can reach: external services, environment variables, shell commands, agent configuration folders, plus any dependencies with known advisories. Claim this listing to get one. How the security profile works

22 of 22 tools

Documented tools (22)

From project documentation. A server handshake does not verify each tool’s description or behavior.

analyze_failure_event

Rank likely instances, source files, and signals for a specific failureevent

analyze_failures

Focus on one grouped failure and return log plus waveform context

build_kdb

Auto-build a Verdi KDB from the parsed compile log (vericom + elabcom). Use when the simulator is Xcelium (xrun) and the NPI backend reports no KDB. Output is cached under TRACEWEAVECACHEDIR (default ~/.cache/traceweave/kdb/<hash>/); cache hits skip re-running Verdi. With TRACEWEAVENPIEXECUTION=lsf,

build_tb_hierarchy

Stream compile evidence and build the full testbench hierarchy server-side without retaining raw source bodies; return a slim payload (project, stats, depth-2 tree skeleton, interfaces, ambiguousbasenames, buildmetrics, hierarchyhandle). For split VCS flows, pass ordered supplementarycompilelogs onc

diff_sim_failure_results

Compare two simulation runs by paths or by basesnapshotid / newsnapshotid. If only newlogpath is supplied after an earlier parsesimlog of the same path, TraceWeave uses the previous parsed snapshot as the baseline.

explain_signal_driver

Trace a waveform signal back to likely RTL driver logic

find_signal_loads

List the consumers (fanout) of a signal — module-input ports, RHS uses, always-block sensitivity

get_diagnostic_snapshot

Read-only summary of cached session data and suggested next calls; mirrors parsesimlog's protocolsymptomhint so a scoreboard failure surfaces the protocol-health pointer at session start

get_error_context

Extract raw log context around a specific line

get_formal_paths

Bounded, tool-neutral discovery of local formal projects, role-labeled logs, and exported VCD/FSDB files. formaltool accepts auto or jaspergold; optional explicit artifact paths must remain inside formalroot. The coverage receipt describes discovery completeness only, and the result intentionally co

get_signal_at_time

Query a signal value at a specific timestamp

get_signal_transitions

Retrieve transitions for a signal over the strict closed interval [starttimeps, endtimeps]; FSDB and VCD never mix an earlier timestamp into transitions. The last value-change strictly before the window is exposed separately as predecessor, which clocked samplers use to classify a transition at the

get_signals_around_time

Retrieve context around a failure timestamp. transitionsinwindow is a strict closed-window list; prewindowtransitions contains only earlier value changes, capped by extratransitions and ordered chronologically on both FSDB and VCD. Flags a valueatcenter that is a sub-cycle transient (a combinational

get_signals_by_cycle

Sample signals cycle-by-cycle on a clock edge

get_sim_paths

Discover compile logs, sim logs, waveforms, simulator, and cases. Optional explicit simlog / wavefile / compilelog overrides win over auto-discovery; omitted fields are still discovered (anchored at the simlog/wavefile directory)

get_waveform_summary

Return waveform metadata. Includes a time-scale self-check: scaleunit (the scale read from the waveform header, e.g. 100fs/1ps/1ns; unknown when unreadable) and scalefspertick — all timestamps in tool output are real picoseconds converted with this factor, never raw file ticks. When the scale is unr

parse_sim_log

Parse and normalize runtime failures into grouped summaries and failureevents; also returns logsnapshotid so same-path reruns can be compared after the simulator overwrites the log. Its candidatepreviouslogs are evidence-backed simulation siblings selected from bounded head/tail samples; compile/ela

recommend_failure_debug_next_steps

Return the default next debug target

scan_structural_risks

Scan compiled RTL/TB sources for structural risk patterns in a lock-free cancellable worker; returns eligiblefilecount, filesscanned, coveragestatus, and coveragewarnings so zero or partial source coverage cannot be mistaken for a clean scan

search_signals

Resolve full hierarchical signal paths. keyword accepts a single string or a list of keywords (max 16) — pass a list to batch several lookups in one call (one result entry per keyword, in input order) instead of issuing consecutive single-keyword searches. Each result also carries direction (input/o

trace_signal_path

Find a structural connectivity path between two signals. Trusted NPI evidence wins; otherwise the bounded dual-endpoint Source Graph follows supported IR bindings and combinational dependencies. Only complete coverage can establish notconnected; an inconclusive negative falls through to unsupportedr

trace_x_source

Trace X/Z propagation upstream through trusted NPI -> one bounded Source Graph artifact -> whole-trace Static. Wave locks cover only waveform reads. Any backend switch or proved scope expansion discards the partial chain and restarts at the original signal, so one returned chain never mixes backend

Tool change history

Compared across complete checks of the same configuration. Tools were listed, not invoked. Input-schema changes are not measured here.

No complete tool checks yet.

FAQ

Questions about TraceWeave MCP Server

How do I connect TraceWeave MCP Server to Claude?: The listing records `<TRACEWEAVE_HOME>/.venv/bin/python <TRACEWEAVE_HOME>/server.py` as its setup step. Run it, then follow the repository's instructions for the client configuration; the listing names Claude Code, Cursor as compatible clients.
Is TraceWeave MCP Server free?: The listed licence is MIT. Check the upstream terms for permitted use and commercial requirements; a public repository does not by itself mean the software is free or open source. Connected APIs and hosted services may have separate charges.
What can TraceWeave MCP Server do?: TraceWeave MCP Server documents 22 tools to the agent, including analyze_failure_event, analyze_failures, build_kdb. The descriptions above come from project documentation. A live handshake does not test individual tool behavior.