Atlas / MCP servers / cachly-dev / cachly — experience for coding agents (MCP server)

cachly — experience for coding agents (MCP server)BLOCK

Name: cachly — experience for coding agents (MCP server)
Author: cachly-dev

mcp/cachly-dev/cachly-mcp

MCP server by cachly-dev · ai · ai agents · anthropic · Audited 2026-10-12

⚑ Report / re-scan Source ↗

Solve it once. From then on, your AI knows. Experience for coding agents over MCP, EU-hosted. 124 MCP tools.

Verdict

BLOCK

Grade

Trust score

45 /100

Exposed tools

124 86r · 33w · 5d

Transport

stdio · streamable-http

License

Apache-2.0

Stars

Overview

From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.

Solve it once. From then on, your AI knows. Every fix, every decision and every dead end becomes experience your AI uses at the right moment. You build faster, and your team builds on what you figured out.

[](https://www.npmjs.com/package/@cachly-dev/mcp-server)

@cachly-dev/mcp-server connects cachly to Claude Code, Cursor, GitHub Copilot, Windsurf, Cline, Continue.dev, Zed, Codex CLI and Gemini CLI over MCP. It offers 124 MCP tools. Your AI stores a lesson after a fix, and finds it again when the same problem returns. Your Brain (the store of your lessons) lives on EU servers.

Start without an account

Claude Code

/plugin marketplace add cachly-dev/cachly-mcp
/plugin install cachly-brain@cachly

There is no JSON to write, no path to set and no key to copy. On the first tool call without a key, cachly creates a test Brain on EU servers. The test Brain has 14 days of the Dev tier. After that it runs on the free tier with the same key.

Check it with claude mcp list. You should see plugin:cachly-brain:cachly ... ✔ Connected. claude plugin details shows MCP servers (0) even when the server runs, because it does not count them.

The plugin adds three Claude Code hooks:

Turn on auto-update once: /plugin → Marketplaces → cachly → Enable auto-update. Claude Code leaves it off for marketplaces outside Anthropic's own.

Read from source at commit f62d26ddffcaOBSERVED · 2026-10-12

Connect

Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.

claude-code (npm)

claude mcp add mcp-server -- npx -y @cachly-dev/[email protected]

Exposed tools (124)

86 read · 33 write · 5 destructive. Blast radius: 5 tools can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.

Tool	Risk	Description
`auto_learn_session`	read	Auto-learn from a list of session observations WITHOUT explicit learn_from_attempts calls.
`autopilot`	read	Generate a CLAUDE.md / copilot-instructions.md that makes any AI self-managing forever.
`brain_briefing`	write	Push-based Brain warning: instead of waiting for you to ask, the Brain proactively checks
`brain_changelog`	read	Generate a human-readable Markdown changelog of lessons learned in the last N days.
`brain_collab_pairs`	read	Show the Person↔Person Collaboration Graph for your team (W5).
`brain_confirm_ci`	read	Close the CI feedback loop: tell the Brain whether a CI job passed or failed and which topics
`brain_conflicts`	read	READ-ONLY — list every unresolved belief_conflict (a previously confirmed fix now contradicted by a failure)
`brain_contribute_signal`	read	Contribute a privacy-safe signal to the global Brain commons. Only the topic category, outcome,
`brain_coverage`	read	Knowledge-coverage health score for your codebase — scored 0-100.
`brain_diff`	read	git log for your AI Brain — see exactly what changed since a point in time.
`brain_discover`	read	Search and browse publicly shared Brain snapshots in the cachly marketplace.
`brain_doctor`	read	Check the health of your AI Brain and get actionable recommendations.
`brain_federate`	write	FedBrain Layer 6 — Private org knowledge transfer: copy CKG edges + lessons from a source brain
`brain_file_map`	read	Show what cachly knows about a list of files — experts + related lessons per file.
`brain_from_ci`	write	Bulk-ingest CI run outcomes into the Brain — the brain_from_git equivalent for CI history.
`brain_from_git`	write	Bootstrap brain lessons from git history. Parses commit messages and infers fix/feature/refactor
`brain_get_pref`	read	Read back one or all preferences stored for this Brain instance.
`brain_graph`	read	Export the Causal Knowledge Graph as a 3D-render-ready node/link payload (schema cachly.brain_graph/v1) —
`brain_hygiene`	read	Autonomously sweep and maintain your Brain — flags stale lessons as provisional, archives long-dormant ones,
`brain_import`	write	Import lessons from a publicly shared Brain snapshot into your own Brain instance.
`brain_import_meta`	write	Import k-anonymous meta-lessons from the global Brain commons into your local Brain.
`brain_install`	write	Install a Domain Brain into your local Brain — pulls its curated, high-trust lessons so they surface in
`brain_marketplace`	read	Browse the Domain Brain marketplace — curated, installable packs of high-trust community lessons,
`brain_metrics`	read	Report the three decisive Brain metrics: (1) time-to-first-recall (onboarding friction),
`brain_plan`	read	READ-ONLY — no side effects, no writes, no external network calls.
`brain_portability`	read	Bring your own model, keep your brain: the same memory in every AI editor.
`brain_predict`	read	READ-ONLY — no side effects, no writes, no external network calls.
`brain_predict_failures`	write	Pre-deploy failure prediction with probability percentages. Given a change context (e.g.
`brain_resolve_conflict`	read	Arbitrate a contested topic by picking the winning side. winner=
`brain_search`	read	BM25+ full-text search over ALL brain data: lessons, context entries, session history, CKG nodes, roadmap items.
`brain_seed_starter`	write	Seed a fresh Brain with a curated set of universal, high-value engineering lessons
`brain_service_map`	read	Map everything the Brain knows about a running service or system: who operates it,
`brain_set_pref`	write	Persist a user preference for this Brain instance.
`brain_share`	write	Export a Brain snapshot and create a publicly shareable link that anyone can import.
`brain_share_list`	read	List all Brain snapshots you have previously shared with brain_share.
`brain_unshare`	destructive	Revoke and permanently delete a public Brain share by its share ID.
`brain_watch`	write	Install an ambient git post-commit hook that automatically learns from every commit —
`brain_who_knows`	read	Find who in your team has the most expertise on a given topic.
`cache_delete`	destructive	Permanently delete one or more keys from a running cache instance (uses Redis DEL).
`cache_exists`	read	Check whether one or more keys exist in a running cache instance (uses Redis EXISTS).
`cache_get`	read	Get a value from a running cache instance by key.
`cache_keys`	read	List keys in a cache instance matching an optional glob pattern (e.g.
`cache_lock_acquire`	write	Acquire a distributed lock using Redis SET NX PX (Redlock-lite).
`cache_lock_release`	read	Release a previously acquired distributed lock.
`cache_mget`	read	Retrieve multiple keys in one round-trip using native Redis MGET.
`cache_mset`	write	Set multiple key-value pairs in a single pipeline round-trip.
`cache_org_stats`	read	Show shared cache statistics for an org namespace.
`cache_set`	write	Set a key-value pair in a running cache instance.
`cache_stats`	read	Get real-time stats for a cache instance: memory usage, hit/miss rate, commands/sec,
`cache_stream_get`	read	Retrieve a previously cached stream as an ordered list of string chunks.
`cache_stream_set`	write	Cache a list of string chunks (e.g. LLM token stream) via Redis RPUSH.
`cache_ttl`	read	Get the remaining time-to-live (TTL) of a key in seconds.
`cache_warmup`	read	Pre-warm the semantic cache with a list of prompt/value pairs.
`causal_trace`	read	Root Cause Analysis through memory: given a problem description, traces the causal chain
`ckg_inspect`	read	Inspect the Causal Knowledge Graph (CKG) for a concept. Shows all typed edges (fixes, requires, co-occurs, causes)
`cls_ingest`	read	Continuous Learning Stream (CLS — Layer 5): Ingest learning signals WITHOUT explicit session_end calls.
`cls_install_hooks`	write	READ-ONLY — outputs text only, writes no files, makes no network calls, has no side effects.
`compact_recover`	read	Call FIRST after any context limit hit / compaction. Reconstructs full context from Memory Crystal +
`create_instance`	write	Create a new managed Valkey/Redis cache instance on cachly.dev.
`create_org`	write	Create a new Cachly organization for team collaboration.
`crystal_view`	read	Inspect the current Memory Crystal — the compressed wisdom distilled from all past sessions.
`delete_instance`	destructive	Permanently delete a cache instance. Deprovisions the Kubernetes workload and removes all data.
`detect_namespace`	read	Classify a prompt into one of 5 semantic namespaces using text heuristics.
`fedbrain_confirm`	read	Confirm that a syndicated lesson from the global commons worked for you.
`fedbrain_contribute`	read	FedBrain (Layer 6): Contribute a lesson to the global Knowledge Commons with a cryptographic
`fedbrain_search`	read	FedBrain context-weighted search: Search the global commons, weighting results by tech-stack similarity.
`fedbrain_status`	read	Show your FedBrain federation status: lessons contributed to global commons, recent confirmations,
`forget_context`	destructive	Delete one or more cached context entries.
`get_api_status`	read	Full diagnostic for your cachly Brain — call this FIRST whenever anything is not working.
`get_connection_string`	read	Get the Redis/Valkey connection string (redis:// URL) for a running instance.
`get_instance`	read	Get full metadata for a specific cache instance: name, tier, status (provisioning / running / paused),
`get_org_plan`	read	Get the current org plan, seat usage, and billing info for an organization.
`global_learn`	read	Store a lesson that applies across ALL your projects (cross-project knowledge).
`global_recall`	read	Read-only retrieval of cross-project lessons stored via global_learn. No side effects.
`import_public_brain`	write	Import community lessons from the Cachly Public Brain for a framework.
`index_project`	read	Index local source files into the cachly semantic cache so AI assistants can use
`invite_member`	read	MUTATION — sends an invite email immediately and cannot be undone via MCP.
`knowledge_decay`	read	Confidence scoring for every lesson in your Brain — because old knowledge rots.
`learn_from_attempts`	write	Store a lesson AFTER any non-trivial task (deploy, debug, fix, decision); future sessions recall it.
`lesson_verified`	read	Report that you RAN the check for a lesson and what came back. This is the only way a lesson gets fresh again —
`list_instances`	read	List all your cachly cache instances with their status and connection details.
`list_orgs`	read	List your Cachly organizations (team/org plans).
`list_remembered`	read	List all cached context entries for this project.
`madc_deliberate`	read	Multi-Agent Deliberation Chamber (MADC — Layer 3): When conflicting lessons exist for a topic,
`memory_consolidate`	read	Cognitive memory consolidation — the weekly garbage collector for your AI Brain.
`memory_crystalize`	read	Compress the last 30-50 sessions and auto-learned lessons into a dense Memory Crystal.
`publish_lesson`	write	Publish a lesson to the Cachly Public Brain (anonymized community knowledge base).
`recall_at`	read	Brain Archaeology — see what a lesson looked like at a specific point in time.
`recall_best_solution`	read	Recall the best known solution for a topic from past lessons.
`recall_context`	read	Retrieve previously saved context from the cache.
`recall_feedback`	read	Tell the Brain whether a recalled lesson ACTUALLY SOLVED your problem.
`remember_context`	write	Save context information to the cache so you can recall it later without re-computing.
`roadmap_add`	write	Add a new item to the persistent project roadmap stored in the Brain.
`roadmap_list`	read	List all roadmap items, optionally filtered by status, priority, tag, or milestone.
`roadmap_next`	read	Get the single most important next actionable roadmap item.
`roadmap_update`	write	Update the status, priority, or details of a roadmap item.
`semantic_search`	read	Find cached entries that are semantically similar to a natural-language query.
`session_end`	write	Save a session summary when you finish working.
`session_handoff`	write	Save a detailed handoff for the NEXT chat window / session.
`session_ping`	read	Lightweight checkpoint — call this every ~5 tool calls or whenever you complete a significant step.
`session_start`	write	Single-call session briefing. Call this at the START of every session INSTEAD of multiple separate smart_recall/recall_best_solution calls.
`session_start_summary`	write	Focused session briefing for large brains. Returns only the top-N most relevant lessons for the given focus topic,
`set_cost_per_call`	write	Set the assumed cost per avoided LLM API call (USD) for this instance.
`setup_ai_memory`	read	One-shot setup of the cachly 3-layer AI Memory system for a project.\n\n
`skill_gaps`	read	Show knowledge blind spots in your Brain — domains with unresolved failures,
`smart_recall`	read	Semantically search cached context using natural language. instance_id may be omitted; the configured instance is used — never guess one.
`sync_file_changes`	write	Associate recent file changes with brain knowledge.
`syndicate`	read	Contribute a verified lesson to the GLOBAL Cachly Knowledge Commons —
`syndicate_search`	read	Search the GLOBAL Cachly Knowledge Commons for solutions contributed by the entire community.
`syndicate_stats`	read	Show the health of the global Knowledge Commons: total lessons, total confirms,
`syndicate_trending`	read	Show the TRENDING lessons in the global Knowledge Commons — those with the fastest confirmation velocity
`team_assign_role`	read	Assign a role (admin \| reviewer \| contributor \| viewer) to a team member on a shared brain instance.
`team_audit`	read	View the governance audit log for a shared brain — an immutable trail of who changed roles and who confirmed which lessons, with timestamps.
`team_confirm`	read	Endorse (review-confirm) a team lesson so trusted, human-reviewed knowledge ranks above unreviewed auto-learned entries.
`team_crystallize`	write	Create a Team Crystal — the team-wide, causal counterpart to memory_crystalize.
`team_expertise_map`	read	Full team expertise overview — who knows what, at a glance.
`team_grant_scope`	destructive	Add or remove a team member to/from a named group (sub-team) on a shared brain.
`team_learn`	read	Store a lesson in a shared team brain so all team members benefit.
`team_recall`	read	Recall lessons from a shared team brain, showing who learned what.
`team_roster`	read	Show all team members and their assigned roles on a shared brain instance.
`team_scopes`	read	List team groups and their members, or the groups a specific person belongs to.
`team_synthesize`	write	Team Brain Synthesis — merge multiple contributors\
`team_whoami`	read	Show your own role and capabilities on a shared brain instance.
`trace_dependency`	read	Causal Chain — find all lessons that depend on a given prerequisite.

Trust audit

BLOCKgrade F · trust 45/100 Do not install this without reading the findings. The audit found something that could harm you or your machine.

Layer	What it checks	Result
L0	Provenance & inventory	PASS
L1	Static analysis of the code	FAIL
L2	Instruction surface (what it tells the agent)	PASS
L3	Class-specific surface	WARN
L4	Behavioural (sandbox)	SKIPPED

What the source does

Filesystem: declared (2 observation(s))
Network: declared (8 observation(s))
Shell: declared (6 observation(s))
Dependencies: not all pinned
Secrets in source: none-found

Findings (25)

HIGHCode injection · code.eval_exec · CWE-78, CWE-94, CWE-95

src/bench/mini-redis.ts:73

pipeline(): { get(k: string): unknown; exec(): Promise<Array<[null, string | null]>> } {

Why it matters. evaluates text as code

Fix. remove; use a parser or a dispatch table

HIGHCode injection · code.eval_exec · CWE-78, CWE-94, CWE-95

src/suchstatistik.ts:28

exec(): Promise<unknown>;

Why it matters. evaluates text as code

Fix. remove; use a parser or a dispatch table

HIGHFilesystem / path · fs.credential_store · CWE-22, CWE-59

src/bench/eingaenge-einbetten.ts:336

console.error('  ssh -i ~/.ssh/cachly-deploy -o "ProxyCommand=ssh -i ~/.ssh/cachly-deploy \\');

Why it matters. touches a credential store

HIGHNetwork egress · net.env_exfil · CWE-200, CWE-319

src/bench/eingaenge-einbetten.ts:336

~/.ssh ... curl

Why it matters. reads secrets in the same file that sends data out

MEDIUMInformation disclosure · disclose.log_secret · CWE-209, CWE-532

src/bench/schablonen-stutzen.ts:152

console.log(`Gestrichen wuerden ${(100 * tokenWeg / tokenAlle).toFixed(1)} % der Token`);

MEDIUMInformation disclosure · disclose.log_secret · CWE-209, CWE-532

src/index.ts:4652

console.log(`     CACHLY_JWT=${token.slice(0, 12)}... CACHLY_BRAIN_INSTANCE_ID=${instance.id}`);