Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/robertoshimizu/session-graph/devkg-sparqlnpx skills add robertoshimizu/session-graph --skill devkg-sparqlgit clone --depth 1 https://github.com/robertoshimizu/session-graphWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/robertoshimizu/session-graph/devkg-sparql)<a href="https://agentmods.dev/skills/robertoshimizu/session-graph/devkg-sparql"><img src="https://agentmods.dev/badge/skills/robertoshimizu/session-graph/devkg-sparql.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00068 | $0.08712 |
| Opus 5 | $0.00034 | $0.04356 |
| Sonnet 5 | $0.00014 | $0.01742 |
| Haiku 4.5 | $0.00007 | $0.00871 |
Grade B, and why
devkg-sparql scanned grade B with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
grep -rli "keyword" ~/.claude/projects ~/.pi/agent/sessions ~/.cursor/projects 2>/dev/null | head -20 Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
- Bash(curl:*) How it starts
The opening of the file, as written. The whole thing — 782 lines — stays where its author put it; the contents beside it link to each section on GitHub.
DevKG SPARQL Query Skill
CRITICAL CONTEXT-SAFETY RULE
NEVER READ LARGE SPARQL RESULTS, SESSION FILES, JSONL, LOGS, OR GENERATED ARTIFACTS ALL AT ONCE. Always add LIMIT, select only needed variables, inspect counts first, and summarize. Never dump huge result sets, ID lists, raw JSON, or transcript content into chat.
Query the developer knowledge graph at http://localhost:3030/devkg/sparql via SPARQL. This graph contains extracted knowledge triples, entities, Wikidata links, and session metadata from Claude Code, pi, Codex, Cursor, ChatGPT, DeepSeek, Grok, and Warp sessions.
Content limit: sioc:content on messages is capped at ~2000 characters at ingest. SPARQL is enough to locate sessions and reason lightly from triples + snippets. For full quotes or deep thread reconstruction, normalize hasSourceFile and read the JSONL only when needed.
Retrieval Strategy (read this first)
| User intent | Do this first | Do NOT start with |
|---|---|---|
| "Where / which sessions discussed X?" | Template 5 (topic + intent + provenance) | Label-only Template 6, or grep |
| "What do we know about technology X?" | Template 1 (entity + provenance) | Grep |
| "How does X relate to Y?" | Template 2 | Grep |
| "Find the exact message wording" | Template 5 or 9 → then JSONL only if snippet truncated | Grepping all projects |
Default for session-discovery questions: multi-signal filter (topic and intent terms) on both triple labels and sioc:content, always joining extractedFrom / extractedInSession, ordered by DESC(?created), with LIMIT.
When Fuseki returns provenance hits, do not fall back to grep. Grep only if Fuseki is down or returns 0 rows after a provenance query.
Execution Pattern
Always use this pattern (POST, URL-encoded query, JSON output). Include Fuseki auth when required:
curl -s -X POST 'http://localhost:3030/devkg/sparql' \
-u admin:admin \
-H 'Accept: application/sparql-results+json' \
-H 'Content-Type: application/x-www-form-urlencoded' \
--data-urlencode "query=YOUR_SPARQL_HERE" \
| jq -r '.results.bindings[] | [.var1.value, .var2.value] | @tsv'
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago First seen · 782 lines · 68 tokens per session scan B c6be999f049b
devkg-sparql is a skill published in the GitHub repository robertoshimizu/session-graph (110 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 68 tokens to every session and 8,712 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it B with 2 findings (reads agent configuration directories, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
graph-mutation-plan
Cookbook for composing an applygraphmutations plan — stable entitykey patterns, the canonical label/edge vocabulary, evidence/invalidation/confidence discipline, and a worked example. Load this when building a non-trivial mutation plan.
open-ontologies
AI-native ontology engineering using 50+ MCP tools backed by an in-memory Oxigraph triple store. Build, validate, query, and govern RDF/OWL ontologies with a generate-validate-iterate loop. Use when building ontologies, knowledge graphs, RDF data, SPARQL queries, BORO/4D modeling, SHACL validation, clinical…
ontology-bootstrap
Build a trustworthy first ontology from an empty or near-empty ontology-atlas vault using only Atlas MCP evidence. Use when the user asks to analyze a codebase, bootstrap/fill its ontology, extract product meaning from a repository, or when a requested ontology task finds only starter nodes. Separate observed…
ontology-field-trial
Measure whether Atlas actually works on a repository nobody here knows, by building a vault with a real MCP agent and then handing that vault — without the source — to a second agent who must answer questions from it alone. Use when changing the construction rules, the MCP read/write contract, the bootstrap skill, or…
design-build
Build UI from this repository's primitives and ramps in a deterministic order, then prove the rendered result with the matching instruments and gates.
design-audit
Audit a finished frontend change by measuring rendered rects and computed styles before using screenshots as human-checkable evidence.