An MCP server over a local humor corpus that credits every line it returns
Wellknown found it in public sources; nobody has proven control of it yet. Claiming takes one click if the repository is under your GitHub account, or a small file on your domain otherwise. Verified owners get the badge, 15-minute checks, status alerts, edits that outrank crawled data, and a ranking boost.
Agents can do it too: POST https://wellknown.network/api/v1/claims with {"agent":"humor-mcp","method":"well_known_file"} — machine-readable steps at claim.json, guide at /docs/claim.
Everything here was measured by our prober or read from a registry. Nothing is self-reported.
Attributed to the source that supplied each field. Treated as claims, not facts.
# humor-mcp **An MCP server over a local humor corpus. Every line it returns says who wrote it.** [](https://pypi.org/project/humor-mcp/) [](https://www.python.org/) [](LICENSE) [](NOTICE) Point a model at a body of jokes and it will happily launder them: you get the material back with no idea whose it was, whether you may publish it, or whether you may sell it. This server treats that as the primary problem. Every result carries the credit for the line it came from, material that may not be redistributed is withheld unless you ask for it by name, and the exporter refuses to hand out anything whose licence does not permit it. No database server, no network, no required dependencies — Python stdlib and one SQLite file. It runs the same whether or not the machine that built the corpus is awake. (That holds for the server and every text path. Audio ingest is the one exception: it needs `numpy`, `scipy` and `soundfile`, and is only imported when you actually use it.) ``` packs/<id>/pack.json who made this material, under what licence packs/<id>/lines.jsonl the lines packs/<id>/pairs.jsonl chosen/rejected preference pairs (optional) humor_mcp/cli.py the one entry point; subcommands below server.py the MCP server (stdio, read-only) build.py packs -> humor.db (SQLite + FTS5) import_audio.py audio + timed transcript (measured reactions) import_transcript.py transcript (text reaction markers) import_corpus.py a jokes file audio_reactions.py the laughter/applause detector transcripts.py shared .srt/.vtt/.json parsing paths.py …
Mapped onto the structured taxonomy from declared text and observed tool names. Confidence shown for derived entries.
Every source is kept verbatim. Field changes are logged as events.