Fulcra context
Access your human's personal context data (biometrics, sleep, activity, calendar, location) via the Fulcra.ai Life API and MCP server. Requires human's Fulcra.ai account + OAuth2 consent.
npx -y skills add arc-claw-bot/fulcra-contextAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Access user-consented Fulcra context data including biometrics, sleep, activity, calendar, location, and the Fulcra metric catalog through the hosted MCP server or Fulcra CLI. Use for bounded read/context workflows only; use fulcra-annotations for writes.
SKILL.md
6.7 KB, ~1.5k tokens by cl100k_base, as published. Nobody here has run it
Fulcra Context
Fulcra gives agents user-consented access to personal context such as biometrics, sleep, activity, calendar, location, and metric catalog data.
This ClawHub skill is intentionally docs-first. It does not ship executable helper scripts, install hooks, background jobs, transcript processors, weather enrichment, export tools, or arbitrary CLI wrappers. Use Fulcra's hosted MCP server or the Fulcra CLI directly, and keep every read bounded to the user's current request.
Use the companion fulcra-annotations skill when an agent needs to create annotation definitions or record user-approved events back to Fulcra.
Privacy Boundary
- Ask before reading Fulcra data.
- Read only the metrics or time window needed for the current answer.
- Do not print, log, paste, or forward access tokens, refresh tokens, credential files, raw private records, or direct capability URLs.
- Do not share real calendar, location, notes, transcripts, or identifying context in public channels unless the user explicitly approves that exact disclosure.
- Do not run transcript, third-party enrichment, raw export, or broad library-file workflows from this skill.
- If the user asks for persistent files, exports, dashboards, or public examples, confirm the exact destination and retention plan first. Prefer synthetic fixtures for public artifacts.
Setup
Option 1: Hosted MCP Server
Use Fulcra's hosted MCP server at:
https://mcp.fulcradynamics.com/mcp
Claude Desktop settings:
{
"mcpServers": {
"fulcra_context": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://mcp.fulcradynamics.com/mcp"]
}
}
}
Local stdio server option:
{
"mcpServers": {
"fulcra_context": {
"command": "uvx",
"args": ["fulcra-context-mcp"]
}
}
}
Open source MCP server: https://github.com/fulcradynamics/fulcra-context-mcp
Option 2: Fulcra CLI
Use the Fulcra CLI for repeatable automation:
uv tool run fulcra-api --help
Authenticate once:
uv tool run fulcra-api auth login
Remote/chat auth rules:
- Keep the CLI process running while it polls for approval.
- Send only the short-lived device URL and user code to the intended user.
- Never send token output or credential files.
- Verify completion with a non-token command:
uv tool run fulcra-api user-info
Fulcra accounts can be created through the CLI and include 5 GB of storage free forever. Users who want phone-collected biometrics, location, calendar, and other context can install the Context iOS app, sign in with the same account, and sync data into that storage.
Allowed Read Workflows
Use these bounded read commands as patterns. Inspect uv tool run fulcra-api --help if the CLI version differs.
Metric Catalog
uv tool run fulcra-api catalog
Use this to discover available data types. Do not query every metric unless the user asks for a broad inventory.
Data Updates
uv tool run fulcra-api data-updates "1 day"
uv tool run fulcra-api data-updates "2026-07-01T00:00:00Z" "2026-07-02T00:00:00Z"
When available, use data-updates as an incremental freshness preflight before broad pulls. It reports which data types had records processed in the requested window and which Fulcra files changed. Use that summary to choose the smallest follow-up read, such as get-records, metric-time-series, sleep-stages, calendar-events, apple-workouts, location commands, or file stat/download commands.
data-updates is not the records themselves. Do not infer health, calendar, or location facts from update counts alone, and do not use it as proof that an annotation write succeeded. Pull the relevant records and verify through the specific read surface.
Recent Heart Rate
uv tool run fulcra-api get-records HeartRate "2 hours"
Summarize aggregates or notable patterns. Avoid dumping raw samples unless the user explicitly asks.
Sleep Context
uv tool run fulcra-api sleep-stages "12 hours"
uv tool run fulcra-api sleep-cycles "1 week"
Use sleep context to adjust briefings or recovery advice. Make clear when data is missing or stale.
Activity And Workouts
uv tool run fulcra-api apple-workouts "1 week"
Keep analysis scoped to activity, recovery, or scheduling questions the user actually asked.
Calendar
uv tool run fulcra-api calendar-events "1 day"
Before using calendar data in a group chat, public artifact, or shared report, ask for explicit permission. Calendar entries may include names, locations, attendees, notes, and meeting links.
Location
uv tool run fulcra-api location-at-time "2026-05-05T12:00:00Z"
Use location only when necessary for the requested context. Do not send coordinates or place history to third-party services from this skill.
Consent Checkpoints
Ask an explicit yes/no question before any workflow that would:
- access calendar or location in a shared chat,
- write local files containing raw Fulcra records,
- create screenshots or charts from private data,
- combine Fulcra data with external services,
- retain data beyond the current session,
- or publish any output outside the user's private workspace.
For public examples, demos, screenshots, videos, tests, and docs, use synthetic data unless the user explicitly approves real data for that exact artifact.
Companion Skills
fulcra-context: read Fulcra context and analyze user-consented health, activity, calendar, location, and catalog data.fulcra-annotations: create annotation definitions and record user-approved events.
Install the companion write-focused skill from ClawHub when needed:
Install the fulcra-annotations skill from ClawHub
Troubleshooting
If a command returns no data:
- Confirm the user has authenticated with
uv tool run fulcra-api user-info. - When live help exposes it, run
uv tool run fulcra-api data-updates <window>to check whether the relevant data types or files changed. - Confirm the user has synced data from the Context app or another source.
- Narrow the query to a known recent time window.
- Report missing or stale data honestly instead of fabricating context.
Links
- Fulcra Platform: https://fulcradynamics.com
- Developer Docs: https://fulcradynamics.github.io/developer-docs/
- Python Client: https://github.com/fulcradynamics/fulcra-api-python
- MCP Server: https://github.com/fulcradynamics/fulcra-context-mcp
What ships with it: 20 files
156.3 KB alongside SKILL.md, 16 of them executable
scripts/
- calendar_vitals_chart.pyruns13.2 KB
- comprehensive_health_dashboard.pyruns25.2 KB
- fulcra_annotations.pyruns4.8 KB
- fulcra_calendar.pyruns763 B
- fulcra_calendar_vitals.pyruns4.1 KB
- fulcra_cli_adapter.pyruns11.9 KB
- fulcra_comprehensive_metrics.pyruns15.7 KB
- fulcra_data_service.pyruns24.0 KB
- fulcra_data_watchdog.pyruns3.4 KB
- fulcra_otter_vitals.pyruns7.3 KB
- fulcra_sleep_utils.pyruns8.0 KB
- fulcra_timezone.pyruns5.0 KB
- sleep_chart.pyruns7.9 KB
- third_space_analyzer.pyruns11.0 KB
- third_space_mapper.pyruns3.8 KB
- weather_provider.pyruns2.6 KB
- .clawhubignore35 B
- .gitignore197 B
- README.md2.7 KB
- SECURITY.md4.6 KB
Gives 0 of the 12 instructions most mcp tooling skills give in ~1.5k tokens
Counted across 638 of the 750 authors here whose files we hold, read 2026-08-07
- Create ten complex or independent read-only evaluation questionsin 69 of 638, across 15 files
- Test servers using MCP Inspectorin 61 of 638, across 19 files
- Provide actionable error messages with specific next stepsin 54 of 638, across 12 files
- Prioritize comprehensive API coverage over specific workflows or workflow toolsin 54 of 638, across 12 files
- Use TypeScript and Streamable HTTP for remote servers or clientsin 54 of 638, across 8 files
- Define structured output schemas where possiblein 50 of 638, across 8 files
- Use Zod or Pydantic for input schemasin 47 of 638, across 5 files
- Fetch MCP specification pages with markdown suffixin 46 of 638, across 4 files
- Load framework documentation using WebFetchin 45 of 638, across 3 files
- Verify each evaluation answer independentlyin 45 of 638, across 3 files
- Implement API client with authentication and paginationin 45 of 638, across 3 files
- Define input schemas with validationin 27 of 638, across 9 files
Said here and by no other author read
- ask before reading fulcra data
- read only metrics or windows needed for the request
- use synthetic data for public artifacts
- ask explicit yes or no before specified workflows
- verify completion with a non-token command
- check data-updates before broad pulls
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.