# BioHarbor (pypi · bioharbor)

Run real bioinformatics from your AI agent. Reliable, reproducible, on your own GPUs.

- Trust score: 91/100 (high trust)
- Change this week: +3
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-11-05

## Components

- pypi · `bioharbor`: 91/100 (this document), [markdown](https://verifymcp.io/servers/danielluo2-bioharbor/bioharbor.md), [page](https://verifymcp.io/servers/danielluo2-bioharbor/bioharbor)

## Channel facts

- Registry: `pypi`
- Package: `bioharbor`
- Version: `0.1.6`
- Transport: `stdio`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-11-05.

- **Supply Chain Security**: 99/100
  - No malware found by supply-chain analysis.
  - No known CVEs affecting this package version or its production dependencies.
  - Runs hatchling.build at install time, a recognised build step with no custom scripting around it.
  - 4 of 36 dependencies flagged as unhealthy.
- **Provenance & Transparency**: 97/100
  - Source repository is publicly reachable at the declared URL.
  - Cryptographically verified build provenance (signed, bound to danielLuo2/bioharbor).
  - Clear OSI-approved license (Apache-2.0).
  - Actively maintained (last published 25 days ago).
  - Disclosure check failed: no security disclosure policy was found in the source repository.
- **Schema Quality & AI Usability**: 73/100
  - AI-judged instruction clarity (excellent).
  - Context-footprint check failed: tool/resource definitions use about 1623 tokens (~135/item across 12 items; 12 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 77/100
  - Stability observed for 23 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 100/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 100% of tool parameters carry a description.
  - Structured output schemas are declared (100% of tools); any adoption earns full credit.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - We read all 12 captured tool definition(s), and no name or description among them implies an irreversible operation.
  - An AI judge read all 13 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 100/100
  - Implements a current MCP spec version (2026-07-28).

## Install

### How do I install the BioHarbor MCP server?

BioHarbor runs locally as a PyPI package, launched with uvx bioharbor. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

### Claude

```bash
claude mcp add danielluo2-bioharbor -- uvx bioharbor
```

### Cursor

```json
{
  "mcpServers": {
    "danielluo2-bioharbor": {
      "command": "uvx",
      "args": [
        "bioharbor"
      ]
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "danielluo2-bioharbor": {
      "command": "uvx",
      "args": [
        "bioharbor"
      ]
    }
  }
}
```

### Codex

```bash
codex mcp add danielluo2-bioharbor -- uvx bioharbor
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "danielluo2-bioharbor": {
      "type": "local",
      "command": [
        "uvx",
        "bioharbor"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add danielluo2-bioharbor --command uvx --arg bioharbor
```

### Hermes

```yaml
mcp_servers:
  danielluo2-bioharbor:
    command: "uvx"
    args: ["bioharbor"]
```

### Netclaw

```json
{
  "McpServers": {
    "danielluo2-bioharbor": {
      "Transport": "stdio",
      "Command": "uvx",
      "Arguments": [
        "bioharbor"
      ]
    }
  }
}
```

### Vellum

```bash
assistant mcp add danielluo2-bioharbor -t stdio -c uvx -a bioharbor
```

### Other

```json
{
  "mcpServers": {
    "danielluo2-bioharbor": {
      "command": "uvx",
      "args": [
        "bioharbor"
      ]
    }
  }
}
```

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-11-05 (score 91, +1)

No change was recorded against any check on this day. Stability & Change Management went from 73 to 77. That category is still filling its 30-day observation window: 22 days of observed history at the previous scan, 23 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-11-02 (score 90, +1)

No change was recorded against any check on this day. Stability & Change Management went from 63 to 67. That category is still filling its 30-day observation window: 19 days of observed history at the previous scan, 20 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-10-31 (score 89, +1)

No change was recorded against any check on this day. Stability & Change Management went from 57 to 60. That category is still filling its 30-day observation window: 17 days of observed history at the previous scan, 18 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-10-29 (score 88, +1)

No change was recorded against any check on this day. Stability & Change Management went from 50 to 53. That category is still filling its 30-day observation window: 15 days of observed history at the previous scan, 16 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-10-27 (score 87, +1)

No change was recorded against any check on this day. Stability & Change Management went from 43 to 47. That category is still filling its 30-day observation window: 13 days of observed history at the previous scan, 14 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-10-25 (score 86, +1)

No change was recorded against any check on this day. Stability & Change Management went from 37 to 40. That category is still filling its 30-day observation window: 11 days of observed history at the previous scan, 12 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-10-23 (score 85, +1)

No change was recorded against any check on this day. Stability & Change Management went from 30 to 33. That category is still filling its 30-day observation window: 9 days of observed history at the previous scan, 10 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-10-22 (score 84, +4)

- [functional improvement] Stability: unverified → 0.30

## MCP tools (12)

### `seq_stats` (~76 tokens)

Validate sequences: type, length, GC%, molecular weight.

Reports dna/rna/protein type, length, GC content (nucleotides) or molecular weight
(proteins) and composition. A cheap first step before heavier tools.

Input parameters:

- `sequence` (string, required): A raw sequence or FASTA text (one or more records).

### `translate_sequence` (~88 tokens)

Translate DNA/RNA to protein, in one frame or all six.

Uses the standard genetic code; stop codons appear as '*'.

Input parameters:

- `frame`: Reading frame; negative = reverse complement; 'all' = six frames.
- `sequence` (string, required): A raw sequence or FASTA text (one or more records).
- `to_stop` (boolean): Stop at the first stop codon.

### `find_orfs` (~207 tokens)

Find open reading frames (ATG to stop codon) in DNA or RNA sequences.

Scans all three frames, on both strands unless both_strands is false, with the standard
genetic code. Returns the top_n longest ORFs (strand, 1-based start/end on the forward
strand, length_nt, length_aa, protein) and the exact input lengths; all ORFs are written
to orfs.faa. If no ORF reaches min_aa, the longest shorter one is reported as
longest_below_min. Protein input is rejected; use translate_sequence for a fixed frame.

Input parameters:

- `both_strands` (boolean): Also search the reverse complement.
- `min_aa` (integer): Minimum ORF length in amino acids (1 = any ORF).
- `sequence` (string, required): A raw sequence or FASTA text (one or more records).
- `top_n` (integer): How many of the longest ORFs to return.

### `search_homologs` (~207 tokens)

Find homologous proteins with MMseqs2 (Swiss-Prot, PDB, ...).

Searches a local database (default Swiss-Prot). Returns the best hits per query with
identity, E-value, coverage, organism and description; the full hit table is written
to a TSV file.

Runs as a background job; may return a job_id to poll with get_job.

Input parameters:

- `database` (string): Installed database name (see list_databases).
- `evalue` (number): E-value cutoff.
- `max_hits` (integer): Max hits per query kept in the output file.
- `min_coverage` (number): Minimum alignment coverage of query and target.
- `sensitivity` (number): MMseqs2 -s: 1 fast … 7.5 most sensitive.
- `sequence` (string, required): Protein (or DNA, searched translated) sequence(s), raw or FASTA.
- `top_n` (integer): Hits per query shown inline.

### `predict_structure` (~113 tokens)

Predict 3D protein structure with ESMFold on a GPU.

Returns per-protein mean pLDDT, confidence bands, low-confidence regions and pTM; PDB
files are written to disk (B-factor column = pLDDT).

Runs as a background job; may return a job_id to poll with get_job.

Input parameters:

- `num_recycles` (integer): Recycling iterations; more can help hard targets.
- `sequence` (string, required): Protein sequence(s), raw or FASTA (≤20 records).

### `get_job` (~162 tokens)

Check a background job and get its result once it has finished.

Use this after a slow tool (e.g. predict_structure, search_homologs) returned a
\`job_id` instead of a result. Returns `status` (queued, waiting_gpu, running,
succeeded, failed, cancelled), plus `result` when succeeded or `error` when failed.
If it is still running, call again later, or pass wait_seconds to block briefly.
Read-only: checking a job does not change it.

Input parameters:

- `job_id` (string, required): The `job_id` returned by a slow tool or listed by list_jobs.
- `wait_seconds` (number): Seconds to wait for the job to finish before returning (0 = return the current status immediately).

### `list_jobs` (~115 tokens)

List recent background jobs, newest first.

Use this to find a `job_id` you no longer have, or to see what is queued or running.
Each entry has job_id, tool, status, gpu and duration_s; results are omitted, so call
get_job for a job's result. Read-only.

Input parameters:

- `limit` (integer): Maximum number of jobs to return.
- `status`: Only return jobs in this state: queued, waiting_gpu, running, succeeded, failed or cancelled. Omit for all states.

### `cancel_job` (~118 tokens)

Cancel a background job that has not started running yet.

Only jobs whose status is queued or waiting_gpu can be cancelled; a job that is
already running is not interrupted and will still finish. Returns `cancelled: true`
on success, or `cancelled: false` with a `message` if the job does not exist or has
already started or finished. Use list_jobs to see job states first.

Input parameters:

- `job_id` (string, required): The `job_id` of the job to cancel (from a slow tool or list_jobs).

### `describe_tool` (~115 tokens)

Explain how an analysis tool runs before calling it.

Returns the tool's description, version, full JSON input schema (parameter types,
defaults and limits), whether it needs a GPU, and whether it runs inline or as a
background job. Use it when unsure about a parameter or after an invalid-parameters
error. Read-only; returns a NotFound error for unknown names.

Input parameters:

- `name` (string, required): Name of an analysis tool: seq_stats, translate_sequence, find_orfs, search_homologs or predict_structure.

### `list_databases` (~102 tokens)

List the sequence databases that search_homologs can search.

Call this before search_homologs to pick a `database` value. Returns `installed`
(name, source and sequence count of each ready database) and `available_to_install`
(known databases such as swissprot, pdb and uniref50 that are not installed yet).
Agents cannot install databases; ask the user to run `bioharbor setup-db <name>`.
Read-only.

### `gpu_status` (~94 tokens)

Show live GPU memory and utilisation on this machine.

Returns, per NVIDIA GPU, its index, name, total and free memory (GB) and utilisation
(%), counting other users' processes too. Use it to explain why a GPU job is in
waiting_gpu, or to check capacity before a large predict_structure run. You do not
need it to schedule jobs: BioHarbor picks a GPU itself. Read-only.

### `read_file` (~141 tokens)

Read a text output file written by a BioHarbor tool.

Tool results are compact summaries; use this only when you need the full output,
e.g. all ORFs in orfs.faa or all hits of a homology search. Returns `content`, the
file's `size_bytes`, and `truncated: true` if the file is longer than max_bytes.
Only files inside the BioHarbor workspace can be read. Read-only.

Input parameters:

- `max_bytes` (integer): Return at most this many bytes from the start of the file.
- `path` (string, required): A path exactly as listed in a tool result's `files` list.

## Diagnostics

Captured diagnostic sections: Provenance, Install scripts, Dependencies. The full working is on the page: https://verifymcp.io/servers/danielluo2-bioharbor/bioharbor#diagnostics

## Score history

- 2026-11-05: 91
- 2026-11-04: 90
- 2026-11-03: 90
- 2026-11-02: 90
- 2026-11-01: 89
- 2026-10-31: 89
- 2026-10-30: 88
- 2026-10-29: 88
- 2026-10-28: 87
- 2026-10-27: 87
- 2026-10-26: 86
- 2026-10-25: 86
- 2026-10-24: 85
- 2026-10-23: 85
- 2026-10-22: 84
- 2026-10-21: 80
- 2026-10-20: 80
- 2026-10-19: 80
- 2026-10-18: 80
- 2026-10-17: 80
- 2026-10-16: 80
- 2026-10-15: 80
- 2026-10-14: 80
- 2026-10-13: 80
- 2026-10-12: 49

## Common questions

### What is the BioHarbor MCP server?

BioHarbor is an MCP server listed in the public MCP registry as io.github.danielLuo2/bioharbor. Run real bioinformatics from your AI agent. Reliable, reproducible, on your own GPUs. This page covers its PyPI package (bioharbor).

### Is the BioHarbor MCP server safe to use?

BioHarbor scores 91 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 5 November 2026. Its build provenance is signed and verified. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

### What tools does the BioHarbor MCP server expose?

BioHarbor exposes 12 tools: seq_stats, translate_sequence, find_orfs, search_homologs, predict_structure, and 7 more. Their descriptions and schemas cost roughly 1,538 tokens of context every time the server is loaded.

### Is the BioHarbor MCP server still maintained?

BioHarbor is still listed as active in the MCP registry. We last reached this channel on 5 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

### What licence is the BioHarbor MCP server under?

BioHarbor declares the Apache-2.0 licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.

## Links

- PyPI project: https://pypi.org/project/bioharbor/
- Socket report: https://socket.dev/pypi/package/bioharbor
- Repository: https://github.com/danielLuo2/bioharbor
- Changelog RSS feed: https://verifymcp.io/servers/danielluo2-bioharbor/bioharbor.xml
- Changelog JSON feed: https://verifymcp.io/servers/danielluo2-bioharbor/bioharbor.json
- HTML version of this page: https://verifymcp.io/servers/danielluo2-bioharbor/bioharbor