osidb-mcp
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@osidb-mcpshow me details for CVE-2023-1234"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
osidb-mcp
Python Model Context Protocol (MCP) server for OSIDB, built on osidb-bindings from PyPI. Use it from Cursor, Claude Desktop, or any MCP client over stdio. PyPI: pypi.org/project/osidb-mcp · Homebrew: vdanen/osidb-mcp · Source: github.com/vdanen/osidb-mcp.
Install
Homebrew
brew tap vdanen/osidb-mcp
brew install osidb-mcpAfter the tap is installed, upgrades use brew upgrade osidb-mcp.
PyPI
Published on PyPI as osidb-mcp:
pipx install osidb-mcp
# or
pip install osidb-mcpPrint the installed package version (no OSIDB env or credentials required):
osidb-mcp --version
# or: osidb-mcp -VConfiguration (environment)
Variable | Required | Description |
| yes | OSIDB root URL, e.g. |
| no |
|
| for | Basic auth for token obtain |
| no |
|
| no | Optional extra User-Agent suffix |
| no | (default). until mutation tools exist. |
Kerberos: the process must have a valid ticket (kinit) for the OSIDB HTTP service.
Optional keys forwarded by bindings: BUGZILLA_API_KEY, JIRA_ACCESS_TOKEN, JIRA_API_EMAIL.
Cursor / Claude MCP snippet
{
"mcpServers": {
"osidb": {
"command": "osidb-mcp",
"env": {
"OSIDB_BASE_URL": "https://your-internal-osidb",
"OSIDB_AUTH": "kerberos",
"OSIDB_VERIFY_SSL": "true",
"OSIDB_MCP_ACCESS_MODE": "readonly"
}
}
}
}Tools (read-only)
All MCP tools require a working OSIDB session (env + Kerberos or basic auth). The CLI osidb-mcp --version / -V does not contact OSIDB (see Install). The table lists every registered tool, in the same order as server.py. For longer explanations, example prompts, and limitations, see TOOLS.md. If an LLM agent is calling these tools, read Using with AI agents first.
Tool | Purpose | |
| OSIDB API health / status payload (good connectivity check). | |
| Current authenticated user / profile from | |
| One flaw by CVE id or internal | |
| High-level search: keyword and/or CVE ids, severity ( | |
| Full flaw plus affects and trackers; | |
| Executive rollups: counts by severity and by workflow plus total under shared filters ( | |
| Raw list API: components, nested affects filters ( | |
| Same filter surface as | |
| Full-text search over flaws ( | |
| Rows keyed by affect with | |
| Tracker filings; scope by | |
| Paginated discussion comments for a flaw id. | |
| Paginated external references (URLs, advisory refs, etc.) for a flaw id. | |
| Paginated CVSS score rows (issuer/version/vector) for a flaw id. | |
| Paginated acknowledgments for a flaw id (CVE or | |
| Paginated flaw collaborator labels for a flaw id. | |
| Paginated package version rows for a flaw id. | |
| Single affect by OSIDB uuid; optional field projection and | |
| Single tracker filing by uuid; optional field projection / | |
| Paginated global OSIDB labels ( | |
| Paginated CVSS rows for one affect (by affect uuid); optional allowlisted | |
| Flaws whose flaw-level components intersect | |
| Affect rows by CVE ( | |
| [EXPERIMENTAL] `GET /exploits/api/v1 | v2/report/pending` — pending exploit / IR actions; may 404 if exploits app is off. |
limit (and analogous list limits) are capped at 100 per request unless noted otherwise on a tool.
When to use which
Triage / natural language style:
search_flaws,get_flaw_details,get_cve_summary.Exact OpenAPI filters or rare query keys:
flaws_list/flaws_countwithextra_query(allowlisted keys only).Affect- or tracker-centric views:
affects_list,trackers_list, or the subresource tools under a known flaw id.
Flaw identifiers (CVE vs internal uuid)
OSIDB flaws always have an internal uuid. A cve_id may be missing until one is assigned — that is normal, not “no identifier.”
List/search responses include each flaw’s
uuidin JSON.flaws_list,flaws_search, andsearch_flaws(structured path) also return anidentifier_hintstring for agents.flaw_get/get_flaw_details: when there is no usable CVE string, the tool adds top-levelosidb_flaw_uuid(same value asflaw.uuid) so follow-up calls are obvious.get_flaw_details: ifcve_idis empty, affects and trackers are loaded usingflaw__uuid/affects__flaw__uuidautomatically.affects_list/query_affects: useflaw_uuid/flaw_uuid_into scope rows when there is no CVE.trackers_list: useaffects_flaw_uuid/affects_flaw_uuid_in.flaw_comments_list,flaw_references_list,flaw_cvss_scores_list,flaw_acknowledgments_list,flaw_labels_list,flaw_package_versions_list: theflaw_idargument is the same as forflaw_get— CVE string or internaluuid. If you useinclude_fieldsonflaw_get, includeuuidwhen you still need it downstream.
Analyst examples
Same idea as “search CVEs”: use
search_flawswithkeyword, or combinecve_ids,severity/severities,date_from/date_to, andproduct_modules/product_components.Critical open flaws touching
httpd:search_flawsorflaws_listwithimpact="CRITICAL",workflow_state_infor non-terminal states, andproduct_components=["httpd"]orcomponents_in/affects_ps_componentas your data model requires.Unowned important CVEs for a RHEL major:
search_flawswithowner_isempty=true,severities=["IMPORTANT"], andproduct_modules/product_componentsset to the exact PS strings your OSIDB uses for that major (confirm in your internal docs).Executive rollup:
get_cve_summarywith optional date range and product filters; tunegroup_byif you only need severity or only workflow buckets.
Using with AI agents
These tools return structured JSON (sometimes large). The MCP host (Cursor, Claude Desktop, API client) chooses the LLM — this server cannot select or downgrade a model for you.
Good default: A mid-tier model (e.g. Sonnet-class) is usually enough for reliable tool names, filters, and reading nested flaw / affect / tracker data.
Smaller / cheaper models: Reasonable for narrow tasks (one CVE, a known tool, counts only). Tight prompts help; ambiguous multi-step triage may need more retries or a larger model.
Largest models: Optional when the task is underspecified or you need unusually careful synthesis; for routine read-only chains they are often more than needed.
Saving tokens: Use
include_fields/exclude_fieldswhere supported; preferflaws_countorget_cve_summaryover pulling many full list pages; keeplimitmodest; ask the agent to summarize instead of echoing entire tool payloads unless you are debugging.
Security
See SECURITY.md for the threat model, OWASP-oriented checklist, and access-mode behavior.
Outputs may include embargoed content; treat transcripts and logs according to your data classification policy (especially when using hosted LLMs).
OSIDB_MCP_ACCESS_MODE=readonlyis the only supported value today —readwriteexits at startup. When mutation tools exist later, use two MCP entries (readonly vs write-capable) so writes happen only when you intend them.Never commit
OSIDB_PASSWORD; use IDE env or secret stores.
Development
uv venv .venv && source .venv/bin/activate
uv pip install -e ".[dev]"
python -m osidb_mcp
pytest tests
pip-auditWith Makefile: make install, make test, make audit, or make check (CI-equivalent). make build / make upload for releases (upload requires twine credentials).
Live integration tests: make livetest runs live_tests/ with pytest -vv -s (verbose + uncaptured stderr) when OSIDB_LIVE_TEST=1 and normal OSIDB_* credentials are set. Default assertions use CVE-2014-0160; optional OSIDB_LIVE_MIN_* env vars enforce minimum totals for your instance. Setup, count output, and .gitignore patterns for secrets are documented in live_tests/README.md. Default make test / pytest only run tests/ (offline).
On Debian/Ubuntu (including local containers), install libkrb5-dev before pip install so the gssapi dependency can find krb5-config (Kerberos stack used with osidb-bindings).
License
MIT — see LICENSE.
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Latest Blog Posts
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/vdanen/osidb-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server