Skip to content
OpenAgentSkill
Browsing is open · Catalog captured 2026-11-21 · Automatic updates and account services are paused.
Browser & automation

ceos-quarterly

Use when conducting, scheduling, or reviewing quarterly conversations between managers and direct reports

bradfeld · AI review on file

Installation

Read the source and dependencies before running a command. Copying does not install or execute anything.

npx skills add bradfeld/ceos --skill ceos-quarterly

About this skill


name: ceos-quarterly description: Use when conducting, scheduling, or reviewing quarterly conversations between managers and direct reports file-access: [data/conversations/, templates/quarterly-conversation.md, data/vision.md, data/accountability.md, data/rocks/, data/people/] tools-used: [Read, Write, Glob]

ceos-quarterly

Facilitate the EOS Quarterly Conversation — the formal quarterly check-in between each manager and their direct reports. This is a two-way conversation about alignment, role satisfaction, and obstacles — not a performance review. It follows a specific 5-point agenda.

When to Use

  • "Run quarterly conversation for [person]" or "quarterly check-in with [name]"
  • "Schedule quarterly conversations" or "who needs a quarterly conversation?"
  • "Review quarterly conversations" or "show conversation history for [person]"
  • "Quarterly one-on-one" or "manager check-in"
  • Any discussion about formal quarterly conversations between managers and direct reports

Context

Finding the CEOS Repository

Search upward from the current directory for the .ceos marker file. This file marks the root of the CEOS repository.

If .ceos is not found, stop and tell the user: "Not in a CEOS repository. Clone your CEOS repo and run setup.sh first."

Sync before use: Once you find the CEOS root, run git -C <ceos_root> pull --ff-only --quiet 2>/dev/null to get the latest data from teammates. If it fails (conflict or offline), continue silently with local data.

Key Files

FilePurpose
data/conversations/QUARTER/Conversation files by quarter (e.g., data/conversations/2026-Q1/)
data/vision.mdSource of Core Values (read-only — use ceos-vto to modify)
data/accountability.mdSource of seats, owners, and reporting structure
data/rocks/QUARTER/Rock files for the quarter (read-only — use ceos-rocks to modify)
data/people/People Analyzer evaluations (read-only — use ceos-people to modify)
templates/quarterly-conversation.mdTemplate for new conversation files

Conversation File Format

Each conversation is a markdown file at data/conversations/YYYY-QN/firstname-lastname.md with YAML frontmatter:

person: "Brad Feld"
manager: "Daniel"
quarter: "2026-Q1"
date: "2026-03-15"
core_values_rating: 5       # count of + ratings out of total Core Values
gwc_status: pass             # pass | fail | evaluating
rocks_completion_rate: 80    # percentage (0-100) or null if no Rocks

File naming: firstname-lastname.md — lowercase, hyphenated. Matches the naming convention from ceos-people.

Quarter Format

Quarters follow YYYY-QN format: 2026-Q1, 2026-Q2, 2026-Q3, 2026-Q4.

To determine the current quarter from today's date:

  • Jan-Mar = Q1, Apr-Jun = Q2, Jul-Sep = Q3, Oct-Dec = Q4

The 5-Point Agenda

#SectionFocus
1Core Values AlignmentHow are they living the Core Values?
2GWCDo they still Get it, Want it, have Capacity?
3Rocks ReviewHow did their Rocks go this quarter?
4Role ExpectationsAre roles clear and being met?
5Feedback Both WaysWhat's working? What's not? What's needed?

Key EOS Principles

  • Every direct report gets one conversation per quarter. This is not optional — it's part of the system.
  • Two-way conversation. The direct report should talk as much as the manager. It's a dialogue, not a lecture.
  • Reference, don't re-evaluate. Core Values and GWC reference the People Analyzer results — this isn't a fresh evaluation. Rock completion is discussed, not re-scored.
  • Document everything. The conversation is recorded and kept for future reference. Git history is the audit trail.

Process

Mode: Facilitate

Use when conducting a quarterly conversation with a specific person.

Step 1: Setup

  1. Identify the person and manager. Ask for the person's name if not provided. Ask who is conducting the conversation (the manager).

  2. Determine the quarter. Default to the current quarter. If near a quarter boundary, ask: "This conversation is for which quarter?"

  3. Check for existing conversation. Look for data/conversations/YYYY-QN/firstname-lastname.md.

    • Exists: Ask: "A conversation already exists for [person] in [quarter]. Open it to review, or start a new one?"
    • New: Continue with Step 2.
  4. Gather context. Read these files (silently — don't dump raw data to the user):

    • data/vision.md — extract Core Values list
    • data/accountability.md — find the person's seat(s) and their manager
    • data/rocks/QUARTER/ — find Rocks owned by this person
    • data/people/firstname-lastname.md — load People Analyzer evaluation if it exists
  5. Create the quarter directory if it doesn't exist: data/conversations/YYYY-QN/

Display a brief preparation summary:

Quarterly Conversation — [Person Name]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Manager: [Manager Name]
Quarter: YYYY-QN
Seat: [From accountability.md]
Core Values: [List from vision.md]
Rocks this quarter: [Count] ([List titles])
People Analyzer: [Status from people/ file, or "No evaluation on file"]

Let's walk through the 5-point agenda.

Step 2: Section 1 — Core Values Alignment

Read Core Values from data/vision.md. Display them and ask the manager to discuss how the person is living each one.

If a People Analyzer evaluation exists (data/people/firstname-lastname.md), show the most recent ratings as a reference point:

People Analyzer reference (last evaluated: 2026-01-15):
  Integrity: +
  Innovation: +/-
  Transparency: +

For each Core Value, prompt: "How is [person] living [Core Value]? Rate: +, +/-, or -"

Record ratings in the conversation file. Calculate core_values_rating as the count of + ratings.

Important: This is a reference discussion, not a re-evaluation. If the ratings differ significantly from t

Read the complete source ↗