Agent skill

Call Summarizer

by CALLE-AI in CALLE-AI/awesome-phone-call-agents

Turn a finished CALL-E phone-call transcript into a structured post-call brief with a one-line outcome, a masked summary, extracted action items with owners and due dates, caller sentiment, and a…

MITAuto-check passedProductivity & Automation

Install Call Summarizer

skills CLI
$ npx skills add CALLE-AI/awesome-phone-call-agents --skill call-summarizer -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install CALLE-AI/awesome-phone-call-agents call-summarizer --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/CALLE-AI/awesome-phone-call-agents.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/call-summarizer .claude/skills/call-summarizer && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
call-summarizer
GitHub stars
107
Token cost
~1.9k tokens
SKILL.md length
950 words
Files
8 (incl. scripts, references)
Skills in repo
101
Repo updated
First seen
Licence
MIT

At a glance

Turn a finished CALL-E phone-call transcript into a structured post-call brief with a one-line outcome, a masked summary, extracted action items with owners and due dates, caller sentiment, and a…

  • Works in 4 steps: Collect the call result → Generate the brief locally → Validate the brief → …
  • Tasks that involve Meeting notes and agendas
  • SKILL.md covers When To Use, When Not To Use, Workflow and Output Schema, plus 4 more sections
  • Runs Python scripts from its folder; calls python3 and pip

What it does

Call Summarizer is an agent skill from CALLE-AI/awesome-phone-call-agents. Turn a finished CALL-E phone-call transcript into a structured post-call brief with a one-line outcome, a masked summary, extracted action items with owners and due dates, caller sentiment, and a redacted caller fingerprint. Use after any CALL-E call when an agent or operator needs an actionable, reviewable record of what was said without re-reading the whole transcript or re-playing the recording.

Its SKILL.md is about 1.9k tokens, which your agent loads only when the skill is triggered. The skill folder holds 9 other files, including scripts and reference files (for example `references/call-e-feedback.md`, `references/example-transcript.json` and `references/examples.md`).

It sits in Productivity & Automation, covering Meeting notes and agendas. The repository describes itself as: Portable phone-call Agent Skills, apps, examples, adapters, and scheduler recipes for AI agents. The licence is MIT.

When your agent uses it

  • Tasks that involve Meeting notes and agendas

Example prompts

  • “/call-summarizer”

Requirements

  • Python 3

Workflow steps

4 steps, taken from the step headings in SKILL.md.

  1. Collect the call result
  2. Generate the brief locally
  3. Validate the brief
  4. Review or route

What it can do on your machine

Read from SKILL.md and the folder at commit 38d4118. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships 3 files in scripts/ (Python), which the agent can run.

    Shell commands in SKILL.md call:

    • python3
    • pip

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use pip, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Call Summarizer loads about 1.9k tokens when it runs, and up to ~5.4k if it reads all its reference files. Until then it costs about 104 tokens; SKILL.md has 950 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~104
When it runs · the whole SKILL.md, loaded when a task matches
~1.9k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~5.4k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from CALLE-AI/awesome-phone-call-agents at commit 38d4118, republished under its MIT licence (© CALLE-AI). 950 words, ~1,945 tokens.

Download SKILL.mdSave it as .claude/skills/call-summarizer/SKILL.md (or your agent's skills folder). This skill also uses 7 other files; get the full folder from GitHub.
name
call-summarizer
description
Turn a finished CALL-E phone-call transcript into a structured post-call brief with a one-line outcome, a masked summary, extracted action items with owners and due dates, caller sentiment, and a redacted caller fingerprint. Use after any CALL-E call when an agent or operator needs an actionable, reviewable record of what was said without re-reading the whole transcript or re-playing the recording.
license
MIT

Call Summarizer

Use this skill after a CALL-E call has completed and the agent needs to turn the returned transcript into a compact, actionable post-call record.

call-summarizer is a post-call analysis skill. It takes a CALL-E call result that already contains a transcript, runs locally with no additional phone calls and no network access, and emits a single structured brief: a one-line outcome, a masked summary of the conversation, the action items with owners and due dates, the caller sentiment, and a redacted caller fingerprint for dedup.

It is a good fit for CALL-E's design: the hard part (the call) is already done, and the remaining work (turning a long transcript into something an agent can act on) is pure text analysis that should not require a second provider or a paid summarization API.

When To Use

Use this skill for:

  • turning a completed CALL-E call transcript into a one-page post-call brief
  • extracting action items with owners and due dates from a call
  • surfacing caller sentiment so a follow-up can be triaged correctly
  • producing a masked summary that is safe to log, store, or hand to a human
  • building a redacted caller fingerprint for de-duplicating repeat callers
  • any workflow where the call is done and the record is the deliverable

When Not To Use

Do not use this skill to:

  • place, schedule, or cancel a phone call; it only reads transcripts
  • summarize a call that has no transcript; it will abstain instead of inventing one
  • act on the action items; it reports them, the operator decides whether to execute
  • store PII; every output is masked and the fingerprint is one-way hashed
  • replace a human review for medical, legal, financial, or emergency content
  • run during the call; it is strictly post-call and never affects call behavior

Workflow

1. Collect the call result

Required: a CALL-E call result containing a transcript field (the full dialogue turns between the agent and the callee). The transcript may be plain text or a list of turns; both are handled.

Confirm with the operator that this transcript belongs to a call they authorized and that they want a post-call brief generated. Never run this skill on a transcript whose origin is unknown.

2. Generate the brief locally

Run scripts/summarize_call.py on the transcript. By default it reads from a file path and prints the brief to stdout; it makes no network calls and places no calls.

bash
python3 scripts/summarize_call.py --transcript path/to/transcript.json --out brief.json

The script performs:

  1. Outcome line: a single sentence stating the call result (confirmed, declined, rescheduled, no-answer, voicemail, unknown) using only words that appear in the transcript. The outcome is bound to the callee's latest effective response (agent text never counts as a confirmation), and any contradictory intent — across utterances or within a single utterance (e.g. "Yes, I can't make it") — fails closed to unknown.
  2. Masked summary: a short prose summary with phone numbers, emails, account identifiers, and title-prefixed or cue-introduced personal names replaced by masked tokens. The brief sets masked: "partial" with a masking_scope field documenting exactly which PII classes are tokenized; ordinary personal names without an introduction cue are NOT redacted (the skill uses no NER model and the contract is honest about this boundary).
  3. Action items: each commitment, follow-up, or next step extracted with an owner (the party who said they would do it), a verb, and an optional due date parsed from natural-language time references. Ambiguous items keep owner: unknown rather than guessing.
  4. Sentiment: a coarse label (positive, neutral, negative, mixed) with a short justification span from the transcript. It never reports a sentiment the transcript does not support.
  5. Caller fingerprint: a one-way hash of a stable caller identity input (the masked caller phone number, or an explicit caller_id field if provided). The call_id is deliberately excluded so the same caller produces the same fingerprint across calls, enabling de-duplication without storing PII.
Show full SKILL.md (311 more words)Show less
3. Validate the brief

Run scripts/validate_brief.py to confirm the brief is well-formed before any downstream system consumes it. It checks that every action item has an owner, that masking has no residual raw phone numbers, emails, account identifiers, or personal names, and that the outcome line is non-empty and grounded in the transcript.

4. Review or route

Return the brief to the operator or the calling agent. The skill does not execute any action item; it only reports them. Routing decisions (escalate, follow up, close the ticket) stay with the operator or the host agent.

Output Schema

The brief is a single JSON object:

json
{
  "outcome": "Appointment confirmed for Tuesday 10:00.",
  "summary": "The callee confirmed the appointment and asked for a reminder the day before.",
  "actions": [
    {
      "owner": "agent",
      "verb": "send reminder",
      "due": "2026-09-15",
      "source_span": "I will send a reminder the day before."
    }
  ],
  "sentiment": {
    "label": "positive",
    "justification": "Callee confirmed without hesitation."
  },
  "caller_fingerprint": "sha256:9f2c...",
  "masked": "partial",
  "masking_scope": "phone_numbers emails account_ids title_prefixed_names cue_introduced_names",
  "masking_note": "Structured PII and cued personal names are tokenized. Ordinary uncued names are NOT redacted."
}

Safety Rules

Read references/safety.md for the full safety contract.

  • This skill never places a call and never modifies call state.
  • Every output is partially masked: phone numbers, emails, account IDs, and cued personal names are tokenized. The masked field is "partial" with a masking_scope documenting the boundary; ordinary uncued names are NOT redacted (no NER model).
  • The caller fingerprint is a one-way hash; the raw identity is never stored.
  • Action items are reported, not executed. Medical, legal, financial, and emergency commitments are flagged as category: sensitive and routed to a human rather than auto-dispatched.
  • If the transcript is empty, garbled, or does not support an outcome, the skill abstains with outcome: unknown and an empty actions list. It never invents a plausible outcome.
  • No PII leaves the local process. There is no network call and no third-party summarization API.

Requirements

  • Python 3.9 or newer. The skill uses only the Python standard library, so no pip install is required for the default (no-call) path.
  • A CALL-E call result with a transcript. Live calls are out of scope; see the call-reminder or verify-by-phone skills for placing calls.

Quick Start

bash
# Dry run on the bundled example transcript (no calls, no network).
python3 scripts/summarize_call.py \
  --transcript references/example-transcript.json \
  --out /tmp/brief.json

# Validate the brief.
python3 scripts/validate_brief.py --brief /tmp/brief.json

Examples

See references/examples.md for worked examples on different call types (confirmation, reschedule, no-answer, voicemail) and the expected brief for each.

© CALLE-AI, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 7 other files (scripts, references) in skills/call-summarizer of CALLE-AI/awesome-phone-call-agents.

  • SKILL.md
  • references/call-e-feedback.md
  • references/example-transcript.json
  • references/examples.md
  • references/safety.md
  • scripts/summarize_call.py
  • scripts/test_summarizer.py
  • scripts/validate_brief.py

Open the folder on GitHubat commit 38d4118

Compare with similar skills

Call Summarizer next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Call Summarizer compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Call Summarizer this skillCALLE-AI/awesome-phone-call-agents107—~1.9kAutomated safety check: PassMIT
Meeting Notesoutline/outline41k—~551Automated safety check: PassCustom licence
Management Talkthananon/9arm-skills3.2k—~3.2kAutomated safety check: PassNone
Challenge Baseline ModelAgibotTech/genie_sim1.4k—~2.4kAutomated safety check: PassCustom licence
Handwriting Stand Uplimin112/min-skill454—~2.5kAutomated safety check: PassNone
Daily Journalravila4/claude-adhd-skills158—~2.5kAutomated safety check: PassMIT

Similar skills

  • Meeting Notes

    outline/outline

    Create meeting notes in Outline from a template; use when the user wants an agenda, notes, or a follow-up document for a meeting.

    41k GitHub stars~551 tokensUpdated yesterday
    Productivity & AutomationAuto-check passed
  • Management Talk

    thananon/9arm-skills

    Rewrite engineer-to-engineer content for engineering-org leadership (VPs, directors, PMs, release managers, execs in an engineering-savvy company) and shape it for the channel it is going to — JIRA…

    3.2k GitHub stars~3.2k tokensUpdated 3 mo ago
    Productivity & AutomationAuto-check passed
  • Challenge Baseline Model

    AgibotTech/genie_sim

    Provision and launch the Simulation Challenge baseline inference model end to end: clone the inference code from a given git repo/branch, download the checkpoints from ModelScope into the repo's…

    1.4k GitHub stars~2.4k tokensUpdated 1 mo ago
    Productivity & AutomationAuto-check passed
  • Handwriting Stand Up

    limin112/min-skill

    Turn a screen recording of handwriting (or a photo of handwritten text) into a single-file HTML "continuation" — the recording plays, the ink lifts off the page as solid 3D letters, a short animated…

    454 GitHub stars~2.5k tokensUpdated yesterday
    Productivity & AutomationAuto-check passed
  • Daily Journal

    ravila4/claude-adhd-skills

    Draft, organize, or update development journal entries. An agent skill from ravila4/claude-adhd-skills.

    158 GitHub stars~2.5k tokensUpdated 7 mo ago
    Productivity & AutomationAuto-check passed
  • Granola

    ArtemXTech/claude-code-obsidian-starter

    Query and sync Granola meetings to Obsidian vault. An agent skill from ArtemXTech/claude-code-obsidian-starter.

    225 GitHub stars~905 tokensUpdated 5 days ago
    Productivity & AutomationAuto-check passed

More from CALLE-AI/awesome-phone-call-agents

All 101 skills in this repo
  • Accessible Outing Verifier

    CALLE-AI/awesome-phone-call-agents

    Demonstrates advisory accessibility-planning checks with offline fixtures and a proposed bounded CALL-E workflow; use for exploring unknown or qualified venue claims without making calls.

    107 GitHub stars~1.4k tokensUpdated yesterday
    Auto-check passed
  • Ground Truth Gate

    CALLE-AI/awesome-phone-call-agents

    A skill your agent uses when an agent holds some evidence for a physical-world claim but the evidence is broader, narrower, or older than the exact question asked, and it must first decide whether a…

    107 GitHub stars~3.3k tokensUpdated yesterday
    Auto-check passed
  • Is It Accessible

    CALLE-AI/awesome-phone-call-agents

    Call a venue and ask the accessibility questions that matter to one specific person — step-free entry, hearing loop, guide dogs, quiet hours, changing places — then return a per-need verdict backed…

    107 GitHub stars~4.3k tokensUpdated yesterday
    Auto-check passed
  • Landmark Navigation Assist

    CALLE-AI/awesome-phone-call-agents

    Turns a pre-written, building-level location config into a CALL-E outbound phone-call task that guides a delivery driver through the last few hundred metres to a specific building using landmarks…

    107 GitHub stars~1.6k tokensUpdated yesterday
    Auto-check passed
  • Research Gap Call Verifier

    CALLE-AI/awesome-phone-call-agents

    Turn cited business research into a bounded, approval-gated phone-call plan that asks only unresolved factual questions, then reconcile CALL-E-compatible results without treating voicemail, refusal…

    107 GitHub stars~1.7k tokensUpdated yesterday
    Auto-check passed
  • Structured Outcome Followup Call

    CALLE-AI/awesome-phone-call-agents

    Place a goal-driven CALL-E call that collects specific structured answers, score those answers against a deterministic rubric you supply, and conditionally trigger a follow-up action — all runnable…

    107 GitHub stars~1.4k tokensUpdated yesterday
    Auto-check passed

Questions about Call Summarizer

What does Call Summarizer do?

Turn a finished CALL-E phone-call transcript into a structured post-call brief with a one-line outcome, a masked summary, extracted action items with owners and due dates, caller sentiment, and a…. Call Summarizer is an agent skill from CALLE-AI/awesome-phone-call-agents. Turn a finished CALL-E phone-call transcript into a structured post-call brief with a one-line outcome, a masked summary, extracted action items with owners and due dates, caller sentiment, and a redacted caller fingerprint.

When should I use Call Summarizer?

Call Summarizer fits situations like: tasks that involve Meeting notes and agendas.

How do I install Call Summarizer in Claude Code?

Run `npx skills add CALLE-AI/awesome-phone-call-agents --skill call-summarizer -a claude-code`. Or copy the skill folder (skills/call-summarizer in CALLE-AI/awesome-phone-call-agents) into .claude/skills/call-summarizer in your project. Claude Code loads it when a task matches its description.

How do I install Call Summarizer in Codex?

Run `npx skills add CALLE-AI/awesome-phone-call-agents --skill call-summarizer -a codex`. Or copy the skill folder (skills/call-summarizer in CALLE-AI/awesome-phone-call-agents) into .agents/skills/call-summarizer in your project. Codex loads it when a task matches its description.

Can I use Call Summarizer in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add CALLE-AI/awesome-phone-call-agents --skill call-summarizer -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/call-summarizer, .gemini/skills/call-summarizer, .github/skills/call-summarizer and .opencode/skills/call-summarizer in your project.

What does Call Summarizer need to run?

Going by SKILL.md and its folder, Call Summarizer needs Python for the scripts in its folder and the command-line tools its instructions call (python3 and pip). Our summary lists: Python 3.

Does Call Summarizer access the network?

SKILL.md contains no URLs. Its commands use pip, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Call Summarizer safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does Call Summarizer use?

Call Summarizer is published under the MIT licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Call Summarizer use?

About 1.9k tokens (SKILL.md is roughly 7.8k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 3.5k tokens, read only when the agent opens those files.

What are the alternatives to Call Summarizer?

Skills that share tags, products or a category with Call Summarizer: Meeting Notes (outline/outline, 41k stars), Management Talk (thananon/9arm-skills, 3.2k stars), Challenge Baseline Model (AgibotTech/genie_sim, 1.4k stars) and Handwriting Stand Up (limin112/min-skill, 454 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Call Summarizer?

CALLE-AI (a GitHub organization) maintains it in CALLE-AI/awesome-phone-call-agents, which has 107 GitHub stars. The repository holds 101 skills in this directory. The repository was last updated on October 10, 2026.

Source: CALLE-AI/awesome-phone-call-agents on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.