io.usefulapi/signl4

What it mainly works on, as our AI model read its tool descriptions: accounts and data in outside web services.

4 of 18 tools are described as changing something (creating, editing, deleting, sending or running).

• signl4_list_teamsreads or searches

  “List the SIGNL4 teams the API key can access: id, name, description, time zone and member ids.”

  Publisher declares: read-only yes · destructive no · open-world yes

  1 input

  • team_idsarray“Limit to these team ids (default: all teams the key can access).”
• signl4_get_teamreads or searches

  “One SIGNL4 team: name, description, time zone, member ids.”

  Publisher declares: read-only yes · destructive no · open-world yes

  1 input, 1 required

  • team_idstringrequired“SIGNL4 team id (a GUID); get it from signl4_list_teams.”
• signl4_list_usersreads or searches

  “List the SIGNL4 users the key can access with name, e-mail, contact addresses, activation state and duty status per team (dutyMode, onDutyTime).”

  Publisher declares: read-only yes · destructive no · open-world yes

  Takes no inputs.

• signl4_get_userreads or searches

  “One SIGNL4 user: name, e-mail, contact addresses, time zone and duty status per team.”

  Publisher declares: read-only yes · destructive no · open-world yes

  1 input, 1 required

  • user_idstringrequired“SIGNL4 user id (a GUID); get it from signl4_list_users.”
• signl4_list_team_usersreads or searches

  “List the users of the given teams (all accessible teams by default) with duty status.”

  Publisher declares: read-only yes · destructive no · open-world yes

  1 input

  • team_idsarray“Limit to these team ids (default: all teams the key can access).”
• signl4_list_categoriesreads or searches

  “List the alert categories of a team (name, color, keywords used to auto-classify alerts).”

  Publisher declares: read-only yes · destructive no · open-world yes

  1 input, 1 required

  • team_idstringrequired“SIGNL4 team id (a GUID); get it from signl4_list_teams.”
• signl4_list_alertsreads or searches

  “Search SIGNL4 alerts (signls), newest first, with status, severity, category and history (created, acknowledged, closed).”

  Publisher declares: read-only yes · destructive no · open-world yes

  9 inputs

  • team_idsarray“Limit to these team ids (default: all teams the key can access).”
  • statusstring“Only alerts in this state (default: all).”
  • created_afterstring“ISO 8601 date-time with a timezone, e.g.”
  • created_beforestring“ISO 8601 date-time with a timezone, e.g.”
  • textstring“Free-text search in title and text.”
  • external_idstring“Only alerts raised with this external id.”
  • category_idsarray“Only alerts of these category ids (see signl4_list_categories).”
  • max_resultsinteger“Alerts per page, 1-100 (default 25).”
  • continuation_tokenstring“The continuation_token of the previous page.”
• signl4_get_alertreads or searches

  “Full detail of one alert: title, text, severity, category, status and delivery status, history (who acknowledged/closed it and when), parameters, attachments and escalations with per-user delivery.”

  Publisher declares: read-only yes · destructive no · open-world yes

  1 input, 1 required

  • alert_idstringrequired“Alert (signl) id; get it from signl4_list_alerts or the id returned by signl4_raise_alert.”
• signl4_get_alert_timelinereads or searches

  “The ordered timeline of an alert: creation, notifications, acknowledgements, annotations, escalations and remote actions.”

  Publisher declares: read-only yes · destructive no · open-world yes

  1 input, 1 required

  • alert_idstringrequired“Alert (signl) id; get it from signl4_list_alerts or the id returned by signl4_raise_alert.”
• signl4_get_alert_reportreads or searches

  “Alert volume and response statistics per team over several time periods: acknowledged, closed and unhandled counts, overall and per category (optionally for one user).”

  Publisher declares: read-only yes · destructive no · open-world yes

  2 inputs

  • team_idsarray“Limit to these team ids (default: all teams the key can access).”
  • user_idstring“Report for this user only.”
• signl4_raise_alertdescribed as changing data

  “WRITE: raise a new alert in a SIGNL4 team.”

  Publisher declares: read-only no · destructive yes · idempotent no · open-world yes

  6 inputs, 3 required

  • team_idstringrequired“SIGNL4 team id (a GUID); get it from signl4_list_teams.”
  • titlestringrequired“Alert title, shown on the pager screen.”
  • textstringrequired“Alert message body.”
  • severitystring“Severity (default: SIGNL4's default for the team).”
  • categorystring“Category name, see signl4_list_categories (default: automatic).”
  • external_idstring“Your own reference for this alert, for later look-up (and for SIGNL4 to group follow-ups).”
• signl4_acknowledge_alertdescribed as changing data

  “WRITE: acknowledge an alert on behalf of a SIGNL4 user (recorded in its history under that user).”

  Publisher declares: read-only no · destructive yes · idempotent yes · open-world yes

  3 inputs, 2 required

  • alert_idstringrequired“Alert (signl) id; get it from signl4_list_alerts or the id returned by signl4_raise_alert.”
  • user_idstringrequired“The SIGNL4 user the acknowledgement is recorded for; see signl4_list_users.”
  • commentstring“Optional comment stored with the acknowledgement.”

Each quoted sentence is the first sentence of the publisher’s own description. The kind of tool and the “changes data” mark are our AI model’s reading of that description, not something we tested: two differently worded runs had to agree, otherwise no label is shown, and nobody has checked them by hand. A tool without the “changes data” mark is described as only reading, searching or calculating, or the two runs did not agree. No tool was called. Read the publisher’s full description before relying on it. “Publisher declares” lists the hints the server itself sends with the tool (MCP annotations: read-only, destructive, idempotent, open-world); they are the publisher's claims, not checked by us. Inputs are the parameters in the tool's own input schema: name, declared type, whether it is required, and the first sentence of its description. Read 2026-10-23.

The registry verifies a package once, when it is published, and does not re-check it. A difference below is about the listing, not a judgement of the publisher.