AI & Machine Learning

bakedin.co

Processes and analyzes academic papers to extract structured claims, compute evidence grades, and build training corpora for language models.

1 endpoint35 known toolsFirst detected June 20, 2026Last detected July 23, 2026

ENDPOINT 1

https://pantry.bakedin.co/corpus/mcp

No auth detected

MCP server metadata

Name: bakedin-corpus
Version: 1.27.0

Capabilities

experimentalpromptsresourcestools

Known tools 35

add_to_corpus

Mark license-verified papers as included in the training corpus.

Inferred read-only

analyze_papers

Extract PICO-structured claims from papers using Bedrock Haiku, then compute GRADE-lite evidence grades deterministically.

Inferred read-only

assemble_corpus

Build a CPT training corpus JSONL from extracted papers.

Inferred read-only

classify_figures

Run Bedrock Haiku classification on unclassified figures.

Inferred read-only

corpus_candidates

Papers that are license-verified and ready for corpus inclusion.

Inferred read-only

corpus_chunk_search

CHUNK-level semantic search — returns the actual paragraphs of body text most relevant to your query, with their paper context.

Inferred read-only

corpus_gap_report

Gap analysis and finish-line cost projection.

Inferred read-only

corpus_hybrid_search

BEST general-purpose search: fuses corpus_keyword_search + corpus_semantic_search via Reciprocal Rank Fusion (RRF).

Inferred read-only

corpus_ingest_graph_edges_jsonl

Ingest course_content_graph edges from S3-hosted JSONL.

Inferred read-only

corpus_ingest_jsonl

Ingest staged papers + chunks from S3-hosted JSONL into corpus_papers + corpus_section_chunks.

Inferred read-only

corpus_keyword_search

Full-text keyword search over the corpus (172K papers).

Inferred read-only

corpus_semantic_search

Meaning-based search over the corpus via pgvector cosine similarity.

Inferred read-only

corpus_state

Current corpus state across all aspects: paper intake, pipeline stages, derived artifacts (claims/SFT/figures/safety/cards), full-text sections, books, regulatory/tribal/underwriting content, quality and relevance distributions, and empirical cost history.

Inferred read-only

corpus_state_trend

Compares the latest corpus snapshot to one from N days ago, showing deltas on headline metrics (papers, abstracts, license-verified, CPT tokens, claims, SFT pairs + source papers, paper sections, books).

Inferred read-only

datacenter_anatomy

Assemble the cross-axis 'Anatomy of a Datacenter': a facility's scale + siting (infra_projects), the energy build-out in its state, the subsidies/cost-shift in its jurisdiction (infra_subsidies), and the corpus papers explaining the underlying constraints.

Inferred read-only

deduplicate

Find and report duplicate papers across different sources (matching DOI or ArXiv ID).

Inferred read-only

discover_domain

Fast metadata-only indexing.

Inferred read-only

domain_coverage

Stats on paper coverage per domain and overall progress toward the 300+ paper corpus target.

Inferred read-only

extract_figures

Download PDFs and extract figures for papers.

Inferred read-only

extract_full_text

Download PDF, extract full text with section parsing (IMRaD), chunk for training.

Inferred read-only

extract_text

Extract text from papers that have abstracts or PDF URLs.

Inferred read-only

figure_catalog

Search and browse extracted figures with filters.

Inferred read-only

figure_detail

Get full metadata for a specific figure by figure_id.

Inferred read-only

generate_sft_from_claims

Generate SFT training Q&A pairs from extracted claims using Bedrock Haiku.

Inferred read-only

harvest_all

Run full harvest across all domains and all sources.

Inferred read-only

harvest_domain

Harvest papers for a specific domain from a source API.

Inferred read-only

ingest_queue_health

Queue-state invariants for the ingest pipeline.

Inferred read-only

paper_detail

Full details for a single paper by paper_id (e.g.

Inferred read-only

paper_registry

Query the paper registry with optional filters.

Inferred read-only

processing_status

Pipeline processing progress across all domains and stages.

Inferred read-only

promote_papers

Verify licenses and promote high-quality discovered papers.

Inferred read-only

reject_paper

Mark a paper as rejected with a reason.

Inferred read-only

rejection_log

View the audit log of papers rejected during harvesting.

Inferred read-only

run_canary

Run canary test: preflight checks, process 5 papers through full pipeline, run all validators.

Inferred read-only

verify_licenses

Batch-verify license status for papers in 'discovered' status.

Inferred read-only

CONNECT WITH APPROVAL

Client installation

Review this server and its permissions before adding it. Secret placeholders must be set locally.

Codex

~/.codex/config.toml

[mcp_servers.bakedin-corpus]
url = "https://pantry.bakedin.co/corpus/mcp"
enabled = true

Claude Code

.mcp.json

{
  "mcpServers": {
    "bakedin-corpus": {
      "type": "http",
      "url": "https://pantry.bakedin.co/corpus/mcp"
    }
  }
}

Claude Desktop

Settings → Connectors → Add custom connector

Name: bakedin-corpus
Remote MCP URL: https://pantry.bakedin.co/corpus/mcp

Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.

Cursor

.cursor/mcp.json

{
  "mcpServers": {
    "bakedin-corpus": {
      "url": "https://pantry.bakedin.co/corpus/mcp"
    }
  }
}

Visual Studio Code

.vscode/mcp.json

Add to Visual Studio Code

{
  "servers": {
    "bakedin-corpus": {
      "type": "http",
      "url": "https://pantry.bakedin.co/corpus/mcp"
    }
  }
}

Generic MCP

Client-specific MCP configuration

{
  "name": "bakedin-corpus",
  "transport": "streamable-http",
  "url": "https://pantry.bakedin.co/corpus/mcp"
}

MCP Inspector

Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.

TRUST AND VERIFICATION EVIDENCE

Trust Data Available

BuiltWith Trust API v2 evidence for bakedin.co was fetched 2026-07-26T06:29:15.385Z and is being refreshed.

Trust status Trusted

bakedin.co is assessed as Trusted: Domain has an established technology history spanning over a year.

Indexed

Evidence is source-attributed and does not guarantee that a third-party server is safe. Risk labels are conservative metadata heuristics.