ai.backskills/backskills
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@ai.backskills/backskills watch the pdf-processing skill and show me what changed in its latest version"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
backskills-mcp
Back skills, because skills change and your agents should know.
A thin, stdlib-only MCP (stdio) client for backskills.ai: watch the skills and MCP servers your agent depends on, get version history and diffs, and ask what to chain next. Ranking is merit-only and never paid.
Use the hosted server (no install)
Streamable HTTP endpoint: https://backskills.ai/mcp
Related MCP server: skillcp
Or run locally
pipx run backskills-mcp # or: uvx backskills-mcpClient config:
{"mcpServers":{"backskills":{"command":"uvx","args":["backskills-mcp"],"env":{"BACKSKILLS_KEY":"bsk_..."}}}}Env vars:
BACKSKILLS_URL(defaulthttps://backskills.ai)BACKSKILLS_KEYoptional; if unset, a free key is registered on first tool call.
Tools
register, search, watch, list_watches, changes, versions, weave_ask, weave_reference, usage.
This package only calls the public backskills.ai HTTP API. MIT licensed.
Available Tools
9 toolschangesC
Diffs for watched skills since an ISO time
Inputs: since.
list_watchesA
List watched skills with current hash and last change
Inputs: no named parameters.
registerC
Get an agent API key (automatic if absent)
Inputs: name.
searchC
Search skills and MCP servers
Inputs: q.
usageC
Your usage and draft pricing preview
Inputs: no named parameters.
versionsC
Version history for a skill id
Inputs: skill_id.
watchC
Watch skills you depend on; optional webhook
Inputs: skills (required), webhook.
weave_askC
Find complementary skills for a goal
Inputs: goal, skill_id.
weave_referenceC
Record that you composed one skill with another
Inputs: goal, to_skill, from_skill.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 9 tools
Most tools target clearly different operations: watch/list_watches/changes form a monitoring trio with distinct meanings, and versions vs changes are separable (history vs diffs). The main soft spot is search vs weave_ask, which both return skills but with different intents (query vs goal-driven complement discovery), and the two weave_ tools require reading descriptions to keep apart.
Mixed conventions: some tools are bare lowercase verbs (register, search, watch), some are bare nouns (changes, versions, usage), and two use a namespace_prefix pattern (weave_ask, weave_reference) alongside an unrelated snake_case list_watches. The names are readable but there is no single predictable pattern.
Nine tools is well within the ideal range and each maps to a concrete need: auth, discovery, monitoring, versioning, composition, and billing. Nothing feels padded or redundant in count.
The lifecycle is largely covered — register, search, watch, list_watches, changes, versions, composition discovery/recording, and usage. The notable gap is no unwatch/stop-watching operation and no way to enumerate recorded weaves, leaving agents unable to reverse actions they can take.
Maintenance
Related MCP Connectors
Audit, pin and diff agent tool lists; preflight and watch remote MCP servers. Free tier; x402.
Search and inspect source-pinned agent skills, plugins and tools through a free public MCP.
Hosted MCP server and Agent Skills for specs, engineering threads, and project memory.
Agent-native catalogue of Baseframe Labs dev tools and MCP servers.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceCatalogues skills and agents into a queryable graph, exposing MCP tools for reference counting, usage scanning, and overlay notes with a built-in visual viewer.1MIT
- AlicenseNot gradedqualityCmaintenanceEnables managing a canonical library of agent skills and MCP servers, syncing them across multiple development harnesses, and adding, importing, or configuring them through MCP tools.44 npmMIT
- AlicenseNot gradedqualityCmaintenanceEnables MCP clients to observe a remote Hermes deployment's agents, sessions, crons, skills, and tools, and to run one-shot remote agent chats via a chat tool.1MIT
- AlicenseNot gradedqualityCmaintenanceEnables any MCP client to watch a running DSH agent and steer it mid-flight, streaming thinking, tool and file notifications while sharing intent and exposing a floor lock that denies the agent's writes while held. It also lets clients interrupt the current turn with an injected alert and delegate reviews through an on-disk inbox.343 npmMIT