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 instructions/vlad-public-code/org.json-kula.valem/claude-mdgit clone --depth 1 https://github.com/vlad-public-code/org.json-kula.valemWrote 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/instructions/vlad-public-code/org.json-kula.valem/claude-md)<a href="https://agentmods.dev/instructions/vlad-public-code/org.json-kula.valem/claude-md"><img src="https://agentmods.dev/badge/instructions/vlad-public-code/org.json-kula.valem/claude-md.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 5.1 | $0.06810 | $0.06810 |
| Opus 5 | $0.03405 | $0.03405 |
| Sonnet 5 | $0.01362 | $0.01362 |
| Haiku 4.5 | $0.00681 | $0.00681 |
Grade A, and why
org.json-kula.valem CLAUDE.md scanned grade A with 0 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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 275 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Valem
Deterministic reactive computation runtime for AI-generated structured data models. A spreadsheet-like computation model for JSON-based agent systems.
Project layout
valem/ ← parent POM (multi-module Maven)
├── valem-core/ ← pure-Java library, no Spring
│ src/main/java/org/json_kula/valem/core/
│ ├── model/ ← ModelSpec and sub-records (deserialized from JSON)
│ ├── graph/ ← DependencyGraph, ModelSpecCompiler, ModelSpecValidator,
│ │ SpecEvolution, ExpressionPathExtractor, CompiledModel
│ ├── state/ ← ModelState, Snapshot, DirtyPropagator, PathConverter
│ ├── engine/ ← ModelRuntime, evaluators, EffectDispatcher,
│ │ EffectiveSchemaBuilder, ExpressionCache,
│ │ SchemaValidator, TestCaseRunner, ModelHistory
│ ├── blob/ ← BlobStore interface + InMemoryBlobStore + FilesystemBlobStore
│ └── llm/ ← LlmClient interface, SpecGenerator (feedback loop),
│ SpecGenerationPrompt (prompt templates)
│
├── valem-service/ ← pure-Java business logic, no Spring
│ src/main/java/org/json_kula/valem/service/
│ ├── ModelService ← orchestrates all use cases (mutate, snapshot, evolve, …)
│ ├── ModelRegistry ← ConcurrentHashMap<id, ModelRuntime>
│ ├── ModelInfo ← DTO: id, version, counts
│ ├── JsonPatchTranslator ← RFC 6902 patch → mutations
│ └── *Exception ← domain exceptions (Not Found, Already Exists, Access Denied, …)
│
├── valem-view/ ← pure-Java view evaluation, no Spring
│ src/main/java/org/json_kula/valem/view/
│ ├── model/ ← ViewDefinition, ViewSpec, ComponentSpec and sub-specs
│ └── engine/ ← ViewEvaluator, EvaluatedView, EvaluatedComponent subtypes
│
├── valem-api/ ← Spring Boot 3 REST + WebSocket layer — HEADLESS library (no main,
│ │ no executable repackaging); embeddable in a host Boot app
│ src/main/java/org/json_kula/valem/api/
│ ├── controller/ ← ModelController, BlobController, ViewController,
│ │ GenerateController, LlmLogController
│ ├── registry/ ← ModelRegistry (Spring wrapper), ModelLoader (ApplicationRunner)
│ ├── persistence/ ← legacy in-api ModelStore (Filesystem/InMemory); the supported
│ │ multi-backend stores live in valem-persistence/* (below)
│ ├── config/ ← StorageConfig (per-concern backend wiring), SecurityConfig,
│ │ ServiceConfig, WebSocketConfig, LlmConfig
│ ├── filter/ ← RateLimitFilter (optional per-IP, off by default)
│ ├── dto/ ← CreateModelResponse, MutationResponse, ValidationErrorResponse
│ ├── llm/ ← AnthropicLlmClient, OpenAiLlmClient, MockLlmClient,
│ │ RecordingLlmClient, LlmInteractionLog, LlmInteractionRecord
│ ├── websocket/ ← ValemWebSocketHandler, ChangeEvent, TokenHandshakeInterceptor
│ └── error/ ← GlobalExceptionHandler (RFC 7807 Problem Detail)
│
├── valem-web/ ← the runnable open web deployable (executable Spring Boot jar): owns
│ ValemWebApplication (main) + the repackaging + the deployer's
│ chosen persistence adapter(s). = valem-api (headless) + memory
│ + filesystem by default; add an adapter jar (or -Pweb-postgres|web-
│ mongo|web-redis|web-s3) for other backends. Builds valem-ui (the
│ management SPA) via frontend-maven-plugin and serves it at / (skip
│ with -Dskip.frontend=true). Also hosts McpHttpController — the MCP
│ surface over Streamable HTTP at /mcp (reuses valem-mcp's McpHttpSession;
│ shared models, valem.api.key gate, valem.mcp.allowed-origins).
│
├── valem-effects-noop/ ← reference effect-kind plugin: ServiceLoader EffectKind + auto-configured
│ EffectExecutor bean; proves the pluggable-effects SPI (drop-in jar, no
│ core/api edits). Template for real custom effect kinds.
│
├── valem-persistence/ ← pluggable storage SPI + adapters (no Spring in the modules)
│ ├── valem-persistence-api/ ← SpecStore, StateStore, ModelStore, CompositeModelStore,
│ │ MutationLogReplay (create-as-you-go), BlobSpooler,
│ │ audit/ (AuditStore, AuditRecord, AuditQuery, DisabledAuditStore)
│ ├── valem-persistence-memory/ ← InMemoryModelStore + InMemoryBlobStore
│ ├── valem-persistence-filesystem/ ← FilesystemModelStore + FilesystemBlobStore
│ ├── valem-persistence-postgres/ ← Postgres Spec/State/Blob stores (spring-jdbc)
│ ├── valem-persistence-mongo/ ← Mongo Spec/State stores + GridFS blob store
│ ├── valem-persistence-redis/ ← Redis Spec/State stores (Lettuce)
│ └── valem-persistence-s3/ ← S3/MinIO blob store (AWS SDK v2)
│
├── valem-console/ ← standalone JSON REPL (no Spring)
│ src/main/java/org/json_kula/valem/console/
│ ├── ConsoleApp
│ └── CommandDispatcher
│
├── valem-mcp/ ← MCP server (Model Context Protocol over stdio, no Spring; same
│ src/main/java/org/json_kula/valem/mcp/ deps as console: service + jackson only)
│ ├── McpServer ← JSON-RPC 2.0 stdio transport + protocol core (initialize/tools/resources/
│ │ logging handshake; negotiates up to 2025-11-25; resources subscribe +
│ │ templates/list; async pair_browser worker w/ progress+cancellation +
│ │ URL-mode elicitation; swappable notification sink for reuse over HTTP)
│ ├── McpHttpSession ← reuses McpServer's core for the Streamable-HTTP endpoint (see valem-web)
│ ├── ToolRegistry ← 24 tools → ModelService: model CRUD (create/mutate/patch_model/
│ │ get_state[paths,depth]/explain/snapshot/restore/get_effective_schema/
│ │ get_audit/verify_audit/upload_blob/download_blob/evolve_spec/…) +
│ │ authoring/verify (validate_spec/eval_expression/test_spec/dry_run) —
│ │ the agent generates, Valem verifies; opt-in mutate traces, outputSchema
│ │ decls, result-size guard, structured errors; + pair_browser in
│ │ remote_with_browser mode (--url --browser)
│ └── ResourceRegistry ← MCP resources: spec-format/jsonata-gotchas/spec-evolution/view-system
│ guides + JSON schemas + bundled example specs + valem://state/{id}
│ subscribable resource (+ examples/state uriTemplates)
│
├── valem-client/ ← thin Java SDK over REST/WS (jackson + JDK java.net.http only;
│ ValemClient, reconnecting subscribe, audit; no engine dep)
│
├── valem-ui/ ← open management/DevTools SPA (Vite+React); built & served at / by
│ valem-web. npm-workspace member alongside valem-view-react.
├── valem-view-react/ ← open EvaluatedView → React renderer library (workspace member)
│
└── clients/valem-sdk-ts/ ← isomorphic TypeScript/JS SDK (fetch + reconnecting WebSocket)
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 · 275 lines · 6,810 tokens per session scan A 6c9593eef9d4
org.json-kula.valem CLAUDE.md is an instructions file published in the GitHub repository vlad-public-code/org.json-kula.valem (0 stars, last pushed 8d ago), licensed Apache-2.0. It adds 6,810 tokens to every session, about $0.0340 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
AI_Agent_MCP_Tool_for_Multi_Docu_PDF_PNG_QR_File_Conversion_Processing_Capability AGENTS.md
AGENTS.md instructions for lovecatisgood-sudo/AI_Agent_MCP_Tool_for_Multi_Docu_PDF_PNG_QR_File_Conversion_Processing_Capability, covering agent guidance, scope, required checks, tool changes and fixtures.
prism AGENTS.md
AGENTS.md instructions for tools-for-agents/prism, covering agents.md — prism, setup, the four verbs, the rules this repo is built on and tests.
mcp-steroid CLAUDE.md
Claude Code instructions for jonnyzzz/mcp-steroid, covering claude.md, agents.md, design philosophy, recursive context lookup (do this before sub-folder work), sub-folder guides and must do.
akm AGENTS.md
AGENTS.md instructions for itlackey/akm, covering akm repo notes, workflow, architecture, tests and test-isolation harness.
json-difference AGENTS.md
Instructions for lukascivil/json-difference, covering general guidelines for working with nx, scaffolding & generators and when to use nxdocs.
autovault CLAUDE.md
Claude Code instructions for autoworks-ai/autovault, covering claude.md, project, common commands, architecture and request flow.