hippocamp
RDF knowledge graph for LLMs — an MCP server that gives any AI agent a structured, queryable memory via six tools: triple, sparql, graph, search, validate, analyze.
Plug it into any project as a persistent brain. Use the built-in ontology for research notes, construction planning, sales pipelines, recipe collections, or any structured knowledge. Auto-installs Claude Code skills and hooks.
On first launch, Hippocamp automatically sets up Claude Code hooks and skills in your project — no manual configuration needed.
Why a graph?
Real test: the same question — "what's needed for electricity?" — asked two ways in a home construction project.
15× faster. 30× less context consumed. Same answer. The LLM learned on its own: "from now on, always search Hippocamp first."
Install
Homebrew (macOS / Linux):
brew install tggo/tap/hippocamp
Direct download — pick your platform from Releases:
# macOS Apple Silicon
curl -L https://github.com/tggo/hippocamp/releases/latest/download/hippocamp_darwin_arm64.tar.gz | tar xz
sudo mv hippocamp /usr/local/bin/
# macOS Intel
curl -L https://github.com/tggo/hippocamp/releases/latest/download/hippocamp_darwin_amd64.tar.gz | tar xz
sudo mv hippocamp /usr/local/bin/
# Linux amd64
curl -L https://github.com/tggo/hippocamp/releases/latest/download/hippocamp_linux_amd64.tar.gz | tar xz
sudo mv hippocamp /usr/local/bin/
Build from source:
git clone https://github.com/tggo/hippocamp
cd hippocamp
make build
Connect to Claude
Add to your MCP config (~/.claude/claude_desktop_config.json or Claude Code settings):
{
"mcpServers": {
"hippocamp": {
"command": "hippocamp",
"args": ["--config", "/path/to/config.yaml"]
}
}
}
Tools
triple — add / remove / list triples
add: {"action":"add","subject":"http://ex.org/Alice","predicate":"http://ex.org/name","object":"Alice","object_type":"literal"}
remove: {"action":"remove","subject":"http://ex.org/Alice","predicate":"http://ex.org/name","object":"http://ex.org/Bob"}
list: {"action":"list","subject":"http://ex.org/Alice"}
sparql — SELECT / ASK / UPDATE
SELECT: {"query": "SELECT ?s ?p ?o WHERE { ?s ?p ?o } LIMIT 10"}
ASK: {"query": "ASK { <http://ex.org/Alice> a <http://ex.org/Person> }"}
UPDATE: {"query": "INSERT DATA { <http://ex.org/Alice> <http://ex.org/name> \"Alice\" }"}
Named: {"query": "SELECT ?name WHERE { ?s <http://ex.org/name> ?name }", "graph": "http://ex.org/g1"}
Returns JSON bindings for SELECT, "true"/"false" for ASK, "ok" for updates.
graph — graph lifecycle, persistence, prefixes
Graph management:
{"action":"list"}
{"action":"create","name":"http://example.org/myGraph"}
{"action":"delete","name":"http://example.org/myGraph"}
{"action":"stats","name":"http://example.org/myGraph"}
{"action":"clear","name":"http://example.org/myGraph"}
Persistence:
{"action":"dump","file":"./backup.trig"}
{"action":"load","file":"./backup.trig"}
Overview and migration:
{"action":"summary"} — compact graph overview (~500 tokens): type counts, top entities, topics, recent decisions
{"action":"migrate"} — apply pending schema migrations (e.g. add provenance defaults to old data)
Namespace prefixes:
{"action":"prefix_add","prefix":"ex","uri":"http://example.org/"}
{"action":"prefix_list"}
{"action":"prefix_remove","prefix":"ex"}
search — semantic keyword search
{"query": "authentication"}
{"query": "Store", "type": "https://hippocamp.dev/ontology#Struct"}
{"query": "middleware", "scope": "project:myapp", "limit": 5}
Searches across rdfs:label, hippo:summary, hippo:alias, hippo:filePath, hippo:signature, hippo:content, hippo:url, hippo:rationale, and subject URIs. Uses field boosting (label matches score highest), word boundary scoring, and score accumulation across predicates.
Use hippo:alias to add synonyms and translations — e.g. Ukrainian labels for English-named resources so search finds them in either language.
With related=true, also returns resources that link TO direct matches via relationship predicates (1-hop graph traversal).
When no results are found, the response includes a hint with the total resource count and suggestions for refining the query.
validate — ontology compliance check
{}
{"scope": "project:house-construction"}
Checks:
- All
rdf:typevalues are from thehippo:namespace - All typed resources have
rdfs:label - All Decisions have
hippo:rationale - Fuzzy type matching: typos like
hippo:Entiyget a suggestion — "did you mean hippo:Entity? (91% similar)" — with fix commands - Schema migrations: warns if a newer schema version is available with
graph action=migratefix
Returns JSON with valid (bool), warnings (array), fixes (array of suggested commands), and stats.
analyze — graph structure analysis
{"action":"god_nodes","limit":5}
{"action":"components"}
{"action":"surprising"}
{"action":"consolidate","limit":10}
{"action":"export_html"}
CLI query mode
For scripts and hooks, run a one-shot search without starting the MCP server:
hippocamp --config config.yaml --query "search terms"
hippocamp --config config.yaml --query "auth" --type "https://hippocamp.dev/ontology#Function" --limit 5
Loads the persisted graph, runs the search, prints pretty-printed JSON, and exits.
Ontology
Hippocamp includes a lightweight RDF ontology (ontology/hippo.ttl) with two layers:
Base layer (any domain):
hippo:Topic— subject areas, themeshippo:Entity— people, orgs, products, APIs, datasetshippo:Note— free-form observationshippo:Source— articles, papers, URLs, bookshippo:Decision— recorded decisions with rationalehippo:Question— open questionshippo:Tag— lightweight labels
Code layer (software projects):
hippo:Project,hippo:Module,hippo:Filehippo:Function,hippo:Struct,hippo:Interface,hippo:Classhippo:Dependency,hippo:Concept
Provenance and quality tracking:
hippo:confidence— float 0.0-1.0 (how certain is this fact?)hippo:provenance—"extracted"/"inferred"/"ambiguous"hippo:source— URI of the agent/tool that produced the resource
Temporal validity:
hippo:validFrom/hippo:validTo— mark facts as time-bounded (invalidated facts stay in graph for history)
The ontology is open-world — extend with your own classes and properties.
Schema migrations
When you upgrade hippocamp, your existing data keeps working (RDF is schema-less). But new features may benefit from enriching old data. The migration system handles this automatically:
- Run
validate— if a migration is available, you'll see:"schema update available — run graph action=migrate to apply 1 pending migration(s)" - Run
graph action=migrate— applies all pending migrations and reports what changed - The graph's schema version is stored as a triple and checked on each validate
Current migrations:
- v2: Adds
hippo:provenance "extracted"andhippo:confidence 1.0to all typed resources that lack them
Migrations are additive — they never delete data. Resources that already have the properties are skipped.
Claude Code integration
Auto-setup
Hippocamp automatically writes hooks and skills to your project's .claude/ directory on first launch. No manual copying needed. On subsequent launches, files are only updated if the binary is newer.
Skill: project-analyze
Ask Claude: "analyze this project and build a knowledge graph". The skill file (.claude/skills/project-analyze.md) is auto-installed and instructs Claude how to scan your project, extract topics, entities, notes, decisions, questions, and sources into the graph.
Hooks
Hooks are auto-installed. You can customize them in .claude/settings.json:
{
"hooks": {
"UserPromptSubmit": [
{ "command": ".claude/hooks/hippocamp-pre-query.sh \"$PROMPT\"" }
],
"PostToolUse": [
{
"command": ".claude/hooks/hippocamp-post-edit.sh \"$TOOL_NAME\" \"$FILE_PATH\"",
"matcher": "Edit|Write"
}
]
}
}
- Pre-query hook: Before each prompt, queries the graph for relevant context
- Post-edit hook: After file edits, tracks which files need re-indexing
Configuration
# config.yaml
store:
default_file: "./data/default.trig" # loaded on startup, saved on SIGINT/SIGTERM
auto_load: true
format: "trig"
prefixes:
ex: "http://example.org/"
schema: "http://schema.org/"
ENV overrides:
Persistence
- Auto-load on startup from
store.default_file(ifauto_load: trueand file exists) - Manual dump via
{"action":"dump","file":"path.trig"} - Auto-save on
SIGINT/SIGTERMif there are unsaved changes
Data is stored in TriG format — a superset of Turtle that preserves named graph boundaries.
Named graphs
All tools accept an optional graph parameter (URI string). Omit it to use the default graph.
# Add to named graph
{"action":"add","graph":"http://ex.org/people","subject":"...","predicate":"...","object":"..."}
# Query named graph
{"query":"SELECT ?s WHERE { ?s ?p ?o }","graph":"http://ex.org/people"}
# Create / delete graphs
{"action":"create","name":"http://ex.org/people"}
{"action":"delete","name":"http://ex.org/people"}
Benchmarks
Memory retrieval benchmarks adapted from LongMemEval (ICLR 2025), ConvoMem (Salesforce), and LoCoMo (Snap Research). Measures R@5 (correct result in top 5) using keyword search only, no embeddings.
Known weakness: adversarial/abstention queries (keyword search can't distinguish "no relevant result" from partial keyword matches).
Run benchmarks:
go test ./internal/tools/ -run TestMemoryBenchmark -v
Visualization
Hippocamp starts an HTTP server on localhost:39322 alongside the MCP server. Open it in a browser for an interactive graph visualization with:
- Nodes colored by type, sized by degree
- Search and type filter
- Click-to-inspect with full property list and edge details
Development
make test # run all tests
make build # build binary
make lint # go vet
make run # build + run with config.yaml
