Skip to main content
Fygment
<- Docs home

Troubleshooting

Live

Troubleshooting

Resolve common Fygment setup, import, MCP, install, support-report, and access issues without leaking sensitive data.

Audience
Users and support operators diagnosing launch-critical workflows.
Updated
2026-05-20

Start with safe, non-sensitive checks. Fygment support reports should capture bounded diagnostics, not package bodies, secrets, raw prompts, or local paths.

Common issues

A docs or setup link redirects to login unexpectedly.

Check: Confirm the route is intended to be public. Public docs, resources, support, privacy, terms, status, and roadmap should not require a session.

Next: Report the route and exact URL through support if a public docs page redirects.

MCP connects but tools return no workspace data.

Check: Confirm OAuth completed for the expected workspace and the token has the needed Fygment scopes.

Next: Reconnect from the MCP setup page, then verify with a simple workspace-visible Skill search.

Import fails for a public source.

Check: Look for candidate selection, validation, unsupported provider, or inaccessible source errors.

Next: Use the report action only after reviewing the disclosure; do not paste tokens or private package content into support.

Support reports

Send safe details only

Good support reports include workspace slug, page or command, timing, non-sensitive error text, and diagnostic id if visible. Do not send API keys, OAuth tokens, private package bodies, secrets, local full paths, stdout, stderr, or raw prompts.

Related docs

  • Set up Fygment MCP

    MCP setup gives AI clients access to workspace-scoped Fygment tools. Hosted clients use OAuth; local clients use guided setup and a local authenticated proxy where needed.

  • Import Skills

    Import turns an existing Skill into a validated Fygment package. Start with one selected package or public source and review the validation result before broader migration.

  • Security and privacy boundaries

    The public launch trust model is metadata-minimizing: keep customer Skill content, package bodies, prompts, secrets, tokens, and raw local paths out of admin, support, analytics, and billing surfaces.