Agent skill

conversation-intelligence

slowingNeeds a keyActs undeclared446 words

Filed under Prospecting and list building.

From devangk003/gtm-agent-skills · 32 skills · 0 · pushed 2026-06-18

What it does when it runs

Extract structured intel from call transcripts and reply corpora — competitor mentions, pricing pushback, feature requests, champion language, blocker signals — and aggregate into pattern-frequency alerts. Use when the user says "extract intel from transcript", "pattern aggregation", "competitor mentions trending", or "feature requests this month.

Read from the skill and the 4 files bundled beside it. A skill’s own description is written to be selected by an agent, so it describes the job and not the dependencies.

Keys and connectors you must supply: AGENTIC_APP_TOKEN
Hosts it reaches: No third-party host appears in the skill or its bundled files.
Tool permissions it declares: No allowed-tools in the frontmatter. It does act, so it runs under whatever permissions your session already grants.
Actions present in the files: shell

Ask about conversation-intelligence

Opens your assistant with this page's verified links already in the prompt.

Is this safe to install?Claude ChatGPT

Adapt it to my stackClaude ChatGPT

What else do I need for it to workClaude ChatGPT

Rather ask a human? Talk to Cheetah

Install it

View source on GitHub ↗

git clone --depth 1 --filter=blob:none --sparse https://github.com/devangk003/gtm-agent-skills.git /tmp/gtm-agent-skills
git -C /tmp/gtm-agent-skills sparse-checkout set "conversation-intelligence"
mkdir -p ~/.claude/skills/conversation-intelligence
cp -R "/tmp/gtm-agent-skills/conversation-intelligence/." ~/.claude/skills/conversation-intelligence/

Picked up without a restart. A project skill of the same name is shadowed by your personal one. For one repository only, swap ~/.claude/skills for .claude/skills. Claude Code docs ↗

The folder is the same in every client that implements the format — 46 of them — so if yours is not above, only the destination changes.

Before you install: this skill will not complete its job on a bare agent. It needs AGENTIC_APP_TOKEN, which you have to obtain separately.

The skill

Source on GitHub ↗

Reproduced in full from devangk003/gtm-agent-skills/blob/6b9a1b0094ffd83d6c02bc03b7ce1610661a1644/conversation-intelligence/SKILL.md, which is licensed MIT (skill frontmatter). 446 words, 6 headings.

Conversation Intelligence

Extract structured intel from unstructured conversation surfaces (Gong/Chorus/Grain/Fathom transcripts, reply text, meeting notes), aggregate into pattern-frequency alerts, and route per pattern class to downstream skills.

When to Use

Gong transcript just dropped — extract intel
Pull patterns from a campaign's reply corpus
Competitor mentions trending this quarter
Feature requests aggregated — prioritize
Champion language patterns across closed-won deals
User says "extract intel from transcript" or "what patterns crossed threshold"

Quick Reference

Pattern class	Detection signal	Default routing
Competitor mention	Named competitor in transcript/reply	→ competitive-intelligence
Pricing pushback	"expensive" / "out of budget" / dollar pushback	→ revenue-forecasting + objection-handling-library
Feature request	"I wish" / "do you have" / named missing feature	→ customer-feedback-analysis + product team
Champion language	Strong-affirmation phrases	→ discovery-call-prep (champion-confirm) + kpi-reporting
Blocker signal	"IT/security/legal/procurement" gating	→ discovery-call-prep (blocker-prep) + pipeline-stages

Concept	Value
Threshold defaults	Competitor: 3/30d / Pricing: 4/30d / Feature: 5/30d
Aggregation window	30d rolling default
Confidence floor	<0.7 → flag for manual review; no auto-route
Verbatim rule	Every pattern must cite actual quote + source timestamp

Procedure

Validate inputs. Pull conversation source (transcript/reply/note). Confirm taxonomy + thresholds loaded. See ${HERMES_SKILL_DIR}/references/pattern-taxonomy.md.
Per-conversation extraction. LLM-backed against 5-class taxonomy. Output: {class, verbatim_quote, source_timestamp, confidence, related_entity}. Confidence <0.7 → manual review.
Per-conversation push. Each extracted pattern → interaction:research with verbatim + source + class. PATCH person/company tags.
Cross-conversation aggregation. Pull all patterns within 30d window. Group by class + entity. Count frequency.
Threshold check + alert. Per class+entity: compare frequency to threshold. Crossed → emit alert routed to downstream skill. See ${HERMES_SKILL_DIR}/references/thresholds.md.
Route per pattern class. Competitor → competitive-intelligence. Pricing → revenue-forecasting. Feature → customer-feedback-analysis. Champion → discovery-call-prep. Blocker → pipeline-stages.
Emit run record. interaction:research with corpus stats + threshold crossings. See ${HERMES_SKILL_DIR}/scripts/push_to_crm.py.

Pitfalls

Paraphrasing instead of verbatim — pattern records MUST contain actual quote
Auto-routing low-confidence extractions — <0.7 → manual review only
Double-counting same conversation across runs — dedup by transcript_id
Treating competitor mentions as objections — "we considered X and rejected them" is positive intel
Aggregation window too narrow — 7d misses trends; 90d includes stale signal

Verification

Every extracted pattern has a verbatim quote + source reference
Threshold-crossing alerts have supporting evidence (per-quote list)
Routing to downstream skills happened for triggered alerts
Aggregation window dates explicit; per-conversation patterns dedup'd

Files bundled with it

These load only when the skill asks for them, so they cost nothing until it runs.

Other skills for the same job

Different authors, same problem. Matched on the words in the skill name, across every library in the catalogue except this one.

conversation-pause-intelligence by louisblythe · 136
personal-strategic-signal-intelligence by ericosiu · 3,449
revenue-intelligence by ericosiu · 3,449
collecting-intelligence by GTM-Strategist · 245
competitive-intelligence-gathering by louisblythe · 136
conversation-ab-testing by louisblythe · 136
conversation-branching by louisblythe · 136
conversation-compliance-auditing by louisblythe · 136

Need help setting it up?

This page tells you what conversation-intelligence does and what it needs. Cheetah builds the agent setup it runs inside: data, CRM, sequencing and the guardrails.

Book a call →

The directory stays free. There is nothing gated behind this.