MCP.Market

Forensic Deepdive MCP server

by Dhevenddra·io.github.Dhevenddra/forensic-deepdive·v0.10.0·2 stars

Persistent code knowledge graph + 9 composite MCP tools for AI coding agents.

B80/100grade B
What users say
No reviews yet
Be the first
Safety scan
B80/100

full report

Adoption
New

2 stars

Reviews

Write one

Nobody has reviewed Forensic Deepdive yet.

If you have run it, two minutes of your experience saves the next person an afternoon.

Forensic Deepdive tools (13, 1 write)

write = sends, deletes, buys or posts

Read from the package source without running it. The installed server may list more.

  • _text
  • add
  • archaeology_tool

    Git-history view of a file (or a symbol's enclosing file): churn, top authors with share, co-change cluster, defect proximity, bus factor, recent commits. Use to answer "who owns this and how risky is it?".

  • context_tool

    One-call 360° view of a symbol: definition + callers + callees + members + siblings + extends/implements + recent commits + dominant author. Use to answer "what does X do and who owns it?".

  • flow_tool

    DFS over CALLS edges from any symbol matching ``entry_point`` up to ``max_depth`` hops. Returns execution-path traces. Use to answer "trace what happens when X is invoked".

  • get_weather
  • impact_tool

    Blast-radius analysis for a symbol. Walks CALLS edges in the given direction (upstream=callers, downstream=callees) up to ``depth`` hops. Returns depth-bucketed lists with confidence labels. Use to answer "what breaks if I change X?".

  • query_toolwrite action

    Run a raw Cypher query against the LadybugDB graph, OR pass ``natural_language`` for a hybrid search (lexical BM25 + structural graph signal + opt-in offline semantic, fused by RRF and shaped to rank implementations above tests). Results carry provenance + confidence; ``retrievers_active``/``degraded`` report which tiers ran. Set ``semantic=True`` to add the ONNX tier (needs the ``[semantic]`` ext

  • recall_insights_tool

    Return prior learnings about *symbol* recorded via record_insight (DEC-019). Newest first, capped at ``limit``. ``since`` (ISO timestamp) filters to insights at or after that time. Use when starting work on a symbol to surface what past sessions learned.

  • record_insight_tool

    Persist one durable learning about this codebase (DEC-019). Use after verifying a hypothesis, fixing a bug, or noticing a cross-session pattern. ``verified_by`` must be one of: "human", "static", "test", "ai". Surfaces back via context() and recall_insights() on future sessions.

  • trace_tool

    Trace one cross-stack feature slice across the `Endpoint` join node (DEC-043). ``downstream`` from a frontend/client symbol walks CALLS_ENDPOINT → Endpoint → HANDLES → handler → CALLS tail; ``upstream`` from a route handler answers "who calls this endpoint". Endpoints with no located handler are surfaced honestly (handler=null). Use to answer "what backend does this component hit?" or "who calls t

  • visualize_tool

    Render a bounded Mermaid diagram of the graph around ``target`` (a symbol or file), or ``central=True`` for the top-``max_nodes`` symbols. Returns a fenced ```mermaid block. Edge dash style encodes confidence (solid=EXTRACTED, dashed=INFERRED, dotted=AMBIGUOUS). Use to answer "show me a diagram of X / how does X connect".

  • x

    Canonicalize an MCP tool name for keying: strip surrounding whitespace and collapse ``. / -`` runs to a single ``_``. Returns ``""`` for an empty/blank name (the caller drops it — no stable id).

Public scan report

scanner v0.1.9 · 2026-09-24 · same rubric, same numbers if you re-run it

1 medium
  • Code scan140 source files scanned20/25
  • –Live reliabilityno gateway calls yet and no remote to proben/a
  • –Tool poisoningtools not inspected (local package is not executed); not countedn/a
  • Auth qualitylocal package, no credentials required12/15
  • Maintenancelast push 39 days ago12/15
  • Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10

Findings (1)

  • mediumeval / new Function usedexec.eval
    forensic_deepdive-0.10.0/experiments/fastcontext/localization_eval.py: …description="deepdive seed localization eval (Arm A)") parser.add_argument("--self…
Overall 80/100. Components that don't apply are left out of the denominator. Any critical finding is an F.RubricAppeal a findingJSON

How Forensic Deepdive compares

Against every graded MCP server, from our own scans and uptime checks. Updated after each scan run.

Safety score
80/100
higher than 60% of 48,255 servers
Tools
13
the typical server has 7
Tools that change things
8%
typical: 0%. Fewer write tools means less an agent can break
GitHub stars
2
more than 75% of 28,688 servers

What its tools let an agent do: 11 read · 1 create · 1 change

Install Forensic Deepdive in Claude Code, Cursor or VS Code

claude mcp add forensic-deepdive -- uvx forensic-deepdive
Add to Cursor

What the publisher says

From the Forensic Deepdive repository's README, as published. We do not edit it. Read it on GitHub

forensic-deepdive

<!-- mcp-name: io.github.Dhevenddra/forensic-deepdive -->

forensic-deepdive gives an AI coding agent forensic understanding of an unfamiliar codebase — a persistent knowledge graph, an MCP server, and five durable markdown artifacts, so your agent stops re-discovering the same architecture from scratch every session.

uv tool install forensic-deepdive && forensic extract /path/to/repo

<!-- regenerate with make demo-gif (scripts/renderdemogif.py — pure Python + Pillow, no vhs/ttyd, DEC-122) -->

Why this is different: every edge and every emitted claim carries a confidence tag — EXTRACTED (deterministic fact) / INFERRED (a heuristic resolved cleanly) / AMBIGUOUS (multiple candidates, shown, not guessed) — so your agent knows what to trust before it acts on it. Nothing else in this category tags confidence at all.

Why not GitNexus / CodeGraphContext / DeepWiki / Sourcegraph

GitNexus is the runaway leader in this category — but its PolyForm Noncommercial license locks every commercial user out. forensic-deepdive is the Apache-2.0 alternative: same persistent-graph-plus-MCP shape, plus honest confidence tags, git archaeology, persistent agent memory, and the 5 markdown artifacts as a fallback for any agent that doesn't speak MCP — with a license your employer's legal team will actually sign off on.

What it produces

forensic-deepdive analyzes any codebase (9 languages, polyglot) and produces:

Shortened. The full README is on GitHub.

Nothing above is checked by us. What we check is on the safety report.

Forensic Deepdive: common questions

Is Forensic Deepdive MCP server safe?
Mostly: it is graded B (80/100). Read the Forensic Deepdive safety report
How do I install Forensic Deepdive?
It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
Does Forensic Deepdive need an API key?
Not as far as the registry entry and our scan can tell: no credentials are declared or required.
Is Forensic Deepdive maintained?
The last commit was 80 days ago (2026-08-16). The latest release is v0.10.0.
What can I use instead of Forensic Deepdive?
Servers from other publishers that do the same job: Codebase Memory MCP server, Neo4j Memory MCP server and State Memory MCP server.
How does Forensic Deepdive compare with other MCP servers?
safety score 80/100 (higher than 60% of 48,255 servers); tools 13 (the typical server has 7); tools that change things 8% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.

Alternatives to Forensic Deepdive

Same job from other publishers: the closest match first, then the best rated.

  • Codebase Memory
    Codebase knowledge graph for AI agents — 162 languages, sub-ms queries, 99% fewer tokens.
    C
  • Neo4j Memory
    MCP Neo4j Knowledge Graph Memory Server
    A
  • State Memory MCP
    Persistent workflow graph server tracking tasks, decisions, spatial entities, and blockers.
    A