MCP server

Stateless Streamable HTTP (JSON-RPC 2.0, protocol 2025-06-18 with fallbacks). The public library needs no key — browsing, evaluating, and installing full skills all work anonymously. A workspace key (upk_…from the dashboard) adds your team's private library and saves.

Claude Code

claude mcp add --transport http uplift https://uplift.page/mcp

any MCP client (JSON config)

{
  "mcpServers": {
    "uplift": { "type": "http", "url": "https://uplift.page/mcp" }
  }
}

with a workspace key (private library + saves)

{
  "mcpServers": {
    "uplift": {
      "type": "http",
      "url": "https://uplift.page/mcp",
      "headers": { "x-uplift-key": "${UPLIFT_KEY}" }
    }
  }
}

install a skill from any repo — no key, no account

curl -s https://uplift.page/mcp -H 'Content-Type: application/json' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"pull_skill","arguments":{"slug":"code-style"}}}' \
  | jq -r '.result.structuredContent.markdown' > skills/code-style/SKILL.md

raw JSON-RPC smoke test

curl -s -X POST https://uplift.page/mcp -H 'Content-Type: application/json' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}' | jq '.result.tools[].name'

Public tools

recommend_stackTech stack + mini prompt + citation for what you describe.

search_promptsFull-text search over the public library.

get_promptFetch one prompt by slug.

list_categoriesList published categories.

convert_promptClean markdown into a tracking-free mini prompt with a removal report.

send_feedbackReport a positive/negative outcome for a prompt.

Skill-library tools

List, install, and score skills — public skills keyless; a workspace key selects your tenant and unlocks the private side. Writes need a paid plan.

workspace_infoThe connected workspace: plan, library counts, hub, month-to-date usage. Key required.

list_skillsList the skill library — public library keyless; a workspace key adds the workspace's own skills.

pull_skillFetch one skill as installable markdown (original SKILL.md or mini prompt) — public skills keyless.

save_skillSave/update a skill in the workspace library: converted, versioned, cited. Key + paid plan.

remove_skillArchive a workspace skill (reversible; history kept). Key + paid plan.

evaluate_skillScore a skill on cleanliness, size, citation, freshness, usage, and outcomes — public skills keyless.

A typical agent flow: recommend_stack with the build intent → adopt the mini prompt → ship → send_feedback with the outcome. For skills: list_skills → pull_skill → write the markdown to skills/<slug>/SKILL.md. The instructions field returned by initialize tells models exactly this.