From crm
Triggered when the user asks what's known about a contact, company, or deal, or wants to update, verify, or manage CRM knowledge. Teaches Claude the progressive retrieval pattern and evidence chain.
How this skill is triggered — by the user, by Claude, or both
Slash command
/crm:memory-layerThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Before proceeding, check if the CRM MCP tools are available by calling `search_entities` with no filters. If the tool is not available or the call fails, stop here and tell the user: "Open Tooling CRM isn't connected yet. Run `/crm-setup` first to get everything installed and connected." Do not proceed with any memory layer operations.
Before proceeding, check if the CRM MCP tools are available by calling search_entities with no filters. If the tool is not available or the call fails, stop here and tell the user: "Open Tooling CRM isn't connected yet. Run /crm-setup first to get everything installed and connected." Do not proceed with any memory layer operations.
You are helping a user interact with Open Tooling CRM's evidence-first memory system.
Raw, immutable evidence. Content-hashed for integrity.
email, call_transcript, meeting_notes, document, noteingest_artifact, get_artifact, list_artifactsTyped, atomic claims extracted from artifacts. Each has a lifecycle.
fact, sentiment, intent, event, metric, relationship, status_changecurrent → superseded or retractedadd_observation, get_observation, list_observations, supersede_observation, retract_observationDerived summaries that cite observations. Always regeneratable.
contact_brief, company_brief, deal_brief, interaction_summarycreate_brief, get_brief, list_briefsExplicit records when observations disagree. Never silently resolved.
create_conflict, get_conflict, list_conflicts, resolve_conflictWhen the user asks "what do we know about X?", follow this pattern:
list_briefs with entity_id to find existing briefslist_observations with entity_id and lifecycle: "current"artifact_id for provenanceget_artifact to retrieve the raw source materialWhen new information replaces old (e.g., budget changed from $100K to $150K):
supersede_observation on the old observation, providing the new observation IDsuperseded lifecycle but is never deletedWhen information turns out to be wrong:
retract_observation with a reasonretracted lifecycleWhen two observations disagree:
resolve_conflict with the resolutionnpx claudepluginhub attri-inc/open-tooling-plugins --plugin crmPersistent intelligence layer that accumulates deal knowledge, competitive intel, product context, and team patterns over time — making every other skill smarter with use. This skill should ALWAYS be checked at the start of any GTM-related task. It reads from and writes to a structured knowledge base so that insights from one conversation carry into the next. Use this skill on every interaction that involves sales, marketing, CS, partnerships, or RevOps context. Also trigger proactively when the user shares deal outcomes, competitive intel, product updates, rep feedback, or customer insights — capture it even if they didn't ask you to.
Syncs monday NoteTaker meeting transcripts to CRM deals by matching participants and appending structured updates with key points, next steps, and commitments.
Conducts multi-source research on customer questions, bugs, account history, or topics from internal docs, CRM, support tickets, chats, and web with attribution and confidence scoring.