agentsclimarketplace

Bi competitive intel

Skill stephenrogan/augmented-csm/skills/pillar-1-book-intelligence/bi-competitive-intel

38 agent skills for the complete CSM operational surface area. Built to the Agent Skills open standard.

Install
npx -y skills add stephenrogan/augmented-csm --skill bi-competitive-intel

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 3 stars3 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

Monitors for signals that a customer may be evaluating alternatives or has been approached by a competitor. Surfaces early warning so the CSM can intervene before a formal evaluation begins. Use when asked to detect competitive threats, monitor for churn to competitors, identify accounts evaluating alternatives, track competitor mentions, build competitive displacement alerts, assess competitive risk across the portfolio, or when any workflow needs to know if a customer is looking at other vendors. Also triggers for questions about win-back, competitor tracking, evaluation detection, or displacement risk.

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

9.7 KB, as published. Nobody here has run it

Competitive Intelligence Monitor

Monitors for signals that customers may be evaluating or being approached by competitors. Part of the Book Intelligence detection layer -- surfaces early warning so the CSM can intervene before a formal evaluation process crystallises.

This is a detection and alerting skill. It identifies competitive signals and surfaces them with evidence. It never contacts the customer about competitive concerns or initiates win-back plays. The response strategy is entirely human territory -- it depends on relationship context, account value, and competitive positioning that the agent cannot assess.

When to Run

  • Continuous: Event-based monitoring of support tickets and call transcripts for keyword matches
  • Scheduled: Weekly scan of behavioural signals (data export activity, engagement patterns)
  • On-demand: When a human requests a competitive risk assessment for a specific account

Competitor Keyword Registry

Maintain a registry of competitor names, products, and evaluation-related terms. The registry should be reviewed and updated quarterly.

Registry categories:

CategoryExamplesNotes
Direct competitors[Company names, product names]Primary competitors in your market
Adjacent competitors[Company names expanding into your space]Products that overlap with part of your value proposition
Evaluation language"alternative," "switch," "migrate," "evaluate," "compare," "looking at options," "RFP"Language that indicates active evaluation regardless of specific competitor
Migration language"data export," "transition plan," "wind down," "offboarding," "contract termination"Language that indicates the evaluation has already progressed to a decision

See references/competitor-registry.md for the full registry and maintenance process.

Core Execution Logic

Step 1: Monitor Direct Mention Signals

Scan incoming data for keyword registry matches:

Support tickets:

  • Scan ticket subject lines and body text for competitor names and evaluation language
  • Weight: subject line matches are stronger signals than body text matches
  • Context matters: "We use [competitor] for [different use case]" is not a risk signal. "[Competitor] offered us a demo for [your use case]" is

Call transcripts:

  • Scan transcripts from Gong, Chorus, or equivalent for competitor mentions
  • Pay attention to the conversation context: who raised the competitor, in what framing
  • Weight: Customer-initiated competitor mention is a stronger signal than CSM-initiated

Email content (if integration available):

  • Scan inbound customer emails for competitor names and evaluation language
  • Weight: Mentions in decision-maker emails are stronger than in end-user emails

Step 2: Monitor Behavioural Signals

Detect evaluation behaviour that does not involve explicit mentions:

SignalDetection MethodStrength
Data export activity spikeProduct analytics: sudden increase in export API calls, bulk downloads, or data extraction feature usageMedium
Admin/API documentation accessProduct analytics: increased visits to API docs, data schema pages, migration guidesMedium
Decreased engagement + no clear causeEngagement score declining with no correlated support issue, product change, or seasonal patternLow-Medium
Concurrent usage of evaluation accountsProduct analytics: new user accounts from the same company domain on a competitor's platform (only if cross-platform tracking is available)High (but rarely available)

Step 3: Score Signal Strength

Strength LevelCriteriaExamples
HighDirect competitor mention in evaluative context by decision-maker"We are evaluating [competitor] for next quarter" in an email from the VP
MediumCompetitor mention without evaluative context, or behavioural signal with clear patternCompetitor mentioned in a support ticket asking about feature comparison; data export spike
LowEngagement decline without explicit competitive indicatorsReduced login frequency with no correlated explanation

Step 4: Correlate with Other Risk Signals

A competitive signal combined with other risk signals amplifies urgency:

  • Competitive signal + health decline = Elevated priority
  • Competitive signal + champion departure = Critical priority
  • Competitive signal + renewal within 90 days = Immediate escalation
  • Competitive signal alone, healthy account = Monitor and prepare

Step 5: Generate Alert

{
  "account_id": "string",
  "account_name": "string",
  "alert_type": "competitive_signal",
  "signal_strength": "high",
  "signal_source": "support_ticket",
  "evidence": {
    "source_type": "support_ticket",
    "source_id": "TICKET-8823",
    "date": "2026-03-08",
    "content_summary": "Customer asked support team how to export all historical data in a format compatible with [Competitor] import tool",
    "contact_role": "Admin (IT Director)"
  },
  "correlated_signals": [
    { "type": "engagement_decline", "detail": "Meeting attendance dropped 40% in last 30 days" }
  ],
  "account_context": {
    "arr": 62000,
    "health_score": 68,
    "days_to_renewal": 112
  },
  "recommended_response_framework": "Direct competitive evaluation detected. Do not lead with competitive positioning. First, investigate whether there is an unmet need driving the evaluation. Address the root cause before defending against the competitor.",
  "urgency": "immediate"
}

Step 6: Track Signal Lifecycle

Each competitive signal has a lifecycle:

  1. New: Signal detected, alert generated
  2. Acknowledged: CSM has seen and accepted the alert
  3. Investigating: CSM is actively gathering more information
  4. Addressed: CSM has engaged the customer and is working to retain
  5. Resolved: Competitive threat neutralised (customer committed to staying)
  6. Lost: Customer has churned to competitor

Track the lifecycle to measure response effectiveness and time-to-response.

Output Format

Competitive threat alert: Individual alert per signal detection (see Step 5 above).

Monthly competitive landscape report:

  • Active threats by competitor (which competitor is winning the most evaluations)
  • Signal source distribution (where are competitive signals detected most frequently)
  • Win/loss tracking (outcomes of competitive engagements)
  • Common competitive triggers (what drives customers to evaluate alternatives)
  • Average response time (how quickly CSMs engage after a competitive signal)

Handoff to Human

All competitive signals surface to the CSM immediately. This is high-judgment territory -- the response depends entirely on relationship context and competitive positioning.

Signal StrengthRoutingUrgency
High (direct mention, evaluative)Alert CSM + notify CS managerImmediate
Medium (behavioural or indirect)Alert CSMSame day
Low (engagement decline only)Include in risk digestNext business day
Any strength + renewal <90 daysAlert CSM + notify CS manager + CRO flagImmediate

See references/human-decision-guide.md for the response framework.

Confidence and Limitations

  • Medium overall confidence. Direct keyword mentions are high-confidence. Behavioural signals are suggestive, not conclusive
  • The agent must present behavioural signals as "possible competitive signal," never "customer is leaving"
  • False positives are common with keyword matching. "We also use [competitor] for [unrelated purpose]" is not a threat signal. Context analysis reduces but does not eliminate false positives
  • Cannot detect competitive evaluations that happen entirely outside observable channels (e.g., the economic buyer receives a competitor pitch at a conference and initiates an evaluation without any digital trace)
  • NLP-based content scanning requires ongoing tuning. New competitors, new product names, and evolving evaluation language require registry updates

Dependencies

Required:

  • Support platform API (ticket content for keyword scanning)
  • Competitor keyword registry (maintained, reviewed quarterly)

Strongly recommended:

  • Call transcript integration (Gong, Chorus) for conversation-level detection
  • Product analytics (for behavioural signals: export activity, documentation access)
  • CRM activity data (engagement patterns)

Optional:

  • Email content integration (for inbound email scanning)
  • Review site monitoring (G2, Capterra -- for public competitive activity)
  • Web tracking (for migration guide or comparison page visits)

Downstream consumers:

  • bi-risk-detector (competitive signals as risk input)
  • bi-account-brief (competitive flags in account context)
  • Human response queue
  • Monthly competitive landscape reporting

References

  • references/competitor-registry.md -- Full keyword registry and maintenance process
  • references/human-decision-guide.md -- Response framework for competitive signals

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.