Skip to main content
Glama

Update Prospect

update_prospect

Set priority=True to push this person's queued sends — email and LinkedIn alike — ahead of everyone else's; it's account-wide and durable, carrying across the sequence's later steps for this person. priority=False clears it. One priority flag per prospect, independent of channel: both channels' derived send order reads it.

Most stage transitions — sent, connected, messaged, replied — stamp automatically from send and reply detection; don't set those by hand. The three reply-outcome rungs above replied — interested/meeting_booked (positive) and not_interested (an explicit decline) — normally stamp from the reply classifier too, so set one by hand only to fix a wrong call or record an off-platform outcome: stage='interested' for clear interest, stage='meeting_booked' when they agree to or book a meeting, stage='not_interested' when they've declined. The mark only raises a rung — a decline can't override a live positive (to lower a false meeting_booked/interested, correct that reply with correct_reply_outcome) — and it counts the person as replied, so to drop someone who never engaged use stage='skipped'. Setting a reply-outcome rung also cancels the person's still-pending queued outreach on both channels — automation stops; messages you queue afterwards still send. A later positive reply lifts not_interested back up on its own.

When you set stage='skipped', pass skip_note with the reason in a few words, so a later reader knows why this person was dropped without digging up the chat — a bare skip records no reason.

To reverse a skip and put the person back in the funnel, set stage to the rung they should resume at — normally stage='pending' to re-queue them from the top. That re-enters them into outreach (a still-skipped prospect is held out of every send) and clears the stale skip note; their flow position is preserved, so the sequence resumes from where they were rather than restarting a step already taken.

To store an address you found for a prospect that reached an email step without one (recovering an email_not_found block), pass email. Resolve that prospect by its LinkedIn handle — channel='linkedin' — since a blank-email prospect has no email to match on; the stamped address makes the step send when you retry it. For a blocked prospect, retry_blocked_enactment(email=...) stamps and retries in one call instead.

To undo an accidental hand-set mark (e.g. a prospect wrongly marked meeting_booked), pass clear_manual_mark=True on channel — it removes the mark and re-derives the stage from actual send/reply activity, the only way to lower past a manual mark (which the mark path deliberately floors demotions at). This differs from correct_reply_outcome, which re-judges a specific reply; use clear_manual_mark when the wrong rung came from a hand mark, not a misjudged reply. Pass it on its own (no stage/data/priority/email). Dict with the updated item details. After a skip, skip_outcome says how many queued sends were cancelled and lists any that had already gone out — a skip can't undo those, so tell the user about them. After an email stamp, profile_note is present only when the person's profile was left as it was because another person in the workspace already carries that address — the campaign row still took it; pass the note on to the user.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
emailNoAn address to store for this prospect (optional). Fills in a prospect that reached an email step without one, so a retried step can send. Lands on the campaign row and on the person's profile, overwriting either's existing address; refused if the address already belongs to another prospect in the account (reconcile the duplicate instead). Leave unset for ordinary stage/priority updates.
stageNoNew stage value (optional) — e.g. 'skipped' to drop the person, 'pending' to reverse a skip and re-queue them.
personYesA handle to the prospect — pass the LinkedIn URL / provider_id / email you already hold; the tool resolves the row within `channel`.
channelYes'linkedin' or 'email' — which funnel the update applies to and which handle on `person` is matched (email against the email column; url/ provider_id against the LinkedIn columns).
agent_idYesID of the agent
priorityNoTrue to prioritize this prospect's queued sends (email and LinkedIn) over everyone else's; False to clear. Account-wide and channel-independent (optional).
skip_noteNoShort free-text reason for a skip, e.g. "off-industry — gold mining, not supplements" (optional). Recorded only when stage='skipped'; ignored otherwise.
data_updatesNoDict to merge into existing data (optional)
linkedin_urlNoA corrected or missing LinkedIn profile URL/slug to bind to THIS prospect in place (optional) — an unverified claim the next send verifies through the profile resolver; any prior verification of the old URL is cleared with it. Refused when the URL already belongs to another of the user's prospects, or when a teammate's prospect verified the same person.
clear_manual_markNoTrue to remove this channel's hand-set reply-outcome mark and re-derive the stage from send/reply activity — the undo for an accidental mark. Pass it alone; combining with stage/data_updates/priority/email is refused.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed3 schema fields changed
    • addedInput schema / properties / agent_id
      Added value: +{
      +  "description": "ID of the agent",
      +  "type": "integer"
      +}
    • removedInput schema / properties / task_id
      Removed value: -{
      -  "description": "ID of the task",
      -  "type": "integer"
      -}
    • changedInput schema / required
      Previous value: -[
      -  "task_id",
      -  "person",
      -  "channel"
      -]New value: +[
      +  "agent_id",
      +  "person",
      +  "channel"
      +]
  2. Changed1 schema field changed
    • changedInput schema / properties / email / description
      Previous value: -"An address to store on the prospect's email column (optional). Fills in a\nprospect that reached an email step without one, so a retried step can send.\nOverwrites this prospect's existing address; refused if the address already\nbelongs to another prospect in the account (reconcile the duplicate instead).\nLeave unset for ordinary stage/priority updates."New value: +"An address to store for this prospect (optional). Fills in a prospect that\nreached an email step without one, so a retried step can send. Lands on the\ncampaign row and on the person's profile, overwriting either's existing\naddress; refused if the address already belongs to another prospect in the\naccount (reconcile the duplicate instead). Leave unset for ordinary\nstage/priority updates."
  3. Changed2 schema fields changed
    • addedInput schema / properties / linkedin_url
      Added value: +{
      +  "anyOf": [
      +    {
      +      "type": "string"
      +    },
      +    {
      +      "type": "null"
      +    }
      +  ],
      +  "default": null,
      +  "description": "A corrected or missing LinkedIn profile URL/slug to bind to THIS\nprospect in place (optional) — an unverified claim the next send verifies\nthrough the profile resolver; any prior verification of the old URL is\ncleared with it. Refused when the URL already belongs to another of the\nuser's prospects, or when a teammate's prospect verified the same person."
      +}
    • changedInput schema / properties / stage / description
      Previous value: -"New stage value (optional, e.g. \"skipped\")"New value: +"New stage value (optional) — e.g. 'skipped' to drop the person, 'pending'\nto reverse a skip and re-queue them."
  4. First observed

TDQS

A5/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With only readOnlyHint=false and destructiveHint=false in annotations, the description carries the behavioral burden and does so thoroughly. It discloses that manual marks only raise a rung, that declines cannot override positive outcomes, that setting a reply outcome cancels pending outreach, that email stamping overwrites existing addresses and can be refused, and that passing linkedin_url clears prior verification.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is long, but the complexity of the tool justifies it and every paragraph covers a distinct decision path with no filler. The summary leads with the core purpose, then logically proceeds through stage semantics, skip/reverse behavior, email stamping, and mark undoing, with a separate returns section.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given 10 parameters, nested objects, no output schema, and minimal annotations, the description is remarkably complete. It covers return value nuances (skip_outcome, profile_note), refusal conditions, cross-channel effects, and the relationship to sibling tools, so an agent has enough context to select and invoke the tool correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Although schema description coverage is 100% (baseline 3), the description substantially enriches parameter semantics: it explains how stage maps per channel, when priority should be set, what skip_note is for, how email resolves blank-email prospects, and how clear_manual_mark differs from correct_reply_outcome. This goes well beyond the individual schema field descriptions.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with a precise verb (update), the resource (a single prospect), and the editable dimensions (stage, data, send priority, contact info). It explicitly routes company updates to update_monitored_company and warns against duplicate creation via track_prospects, so it is easily distinguished from siblings.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Usage guidance is explicit and disambiguated: it names update_monitored_company for companies, retry_blocked_enactment for blocked prospects, correct_reply_outcome for misjudged replies, and clear_manual_mark for accidental hand-set marks. It also states when NOT to use the tool (never re-add a person with changed contact info) and when to set reply outcomes (only to fix or record off-platform).

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources