Ops Recap logo

Ops Recap

Organization
Lifecycle-Innovations-Limited
ops-recap

OPS on-demand: This skill should be used when the user asks to "recap daemon", "tmux marquee", or…

Overview

PublisherLifecycle-Innovations-Limited
Repositoryclaude-ops
Skill nameops-recap
Stars
543
Forks
64
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by Lifecycle-Innovations-Limited on GitHub. Read the source before you install it.

Installation

Install the Ops Recap AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/Lifecycle-Innovations-Limited/claude-ops/tree/main/claude-ops/skills/ops-recap
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/Lifecycle-Innovations-Limited/claude-ops.git /tmp/claude-ops
mkdir -p .claude/skills
cp -r /tmp/claude-ops/claude-ops/skills/ops-recap .claude/skills/ops-recap
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Ops Recap in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Ops Recap on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Ops Recap is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

OPS ► RECAP

Load ops-rules before acting. Public repo (no personal data). Outbound: one draft → one approval → one send. If AskUserQuestion / Workflow are missing, follow Rule 10 in ops-rules (Hermes: numbered options / two-turn Telegram card; delegate_task).

The recap marquee is a launchd-managed daemon that aggregates per-session Claude transcripts (via the recap-capture Stop hook) and per-tool activity (via the recap-tool-activity PostToolUse hook) into a single rolling headline at /tmp/claude-recap-digest. Tmux reads that file every refresh interval via scripts/recap/marquee.sh and scrolls it across status-right.

Files:

PathRole
${CLAUDE_PLUGIN_ROOT}/scripts/recap/daemon.shBackground loop — polls inputs, calls digest.sh when stale
${CLAUDE_PLUGIN_ROOT}/scripts/recap/digest.shSynthesizes one-line digest via claude -p (session default model; OPS_RECAP_MODEL overrides)
${CLAUDE_PLUGIN_ROOT}/scripts/recap/marquee.shTmux-side scroller (called from status-right)
${CLAUDE_PLUGIN_ROOT}/hooks/recap-capture.shStop hook → writes /tmp/claude-recap-<sid>
${CLAUDE_PLUGIN_ROOT}/hooks/recap-tool-activity.shPostToolUse hook → live activity per session
~/Library/LaunchAgents/com.claude-ops.recap-daemon.plistLaunchd agent (macOS)
/tmp/claude-recap-digestRolling one-liner (read by tmux + marquee.sh)
/tmp/claude-recap-daemon.logDaemon log (rotated at 500KB)
/tmp/claude-recap-digest.logLast 50 generated digests, chronological

Subcommand: status (default)

Print a compact health panel:

bash
PLIST=~/Library/LaunchAgents/com.claude-ops.recap-daemon.plist
LABEL=com.claude-ops.recap-daemon
DIGEST=/tmp/claude-recap-digest
DAEMON_LOG=/tmp/claude-recap-daemon.log

# Plist installed?
if [ -f "$PLIST" ]; then echo "✓ plist installed: $PLIST"; else echo "✗ plist missing"; fi

# Launchd loaded?
launchctl print "gui/$(id -u)/${LABEL}" >/dev/null 2>&1 \
  && echo "✓ launchd loaded ($LABEL)" \
  || echo "✗ launchd not loaded — run /ops:recap restart"

# Daemon process alive?
if [ -f /tmp/claude-recap-daemon.pid ]; then
  pid=$(cat /tmp/claude-recap-daemon.pid)
  kill -0 "$pid" 2>/dev/null && echo "✓ daemon alive (pid $pid)" || echo "✗ stale pid file"
else
  echo "✗ no pid file"
fi

# Digest freshness
if [ -f "$DIGEST" ]; then
  age=$(($(date +%s) - $(stat -f %m "$DIGEST" 2>/dev/null || stat -c %Y "$DIGEST" 2>/dev/null || echo 0)))
  echo "✓ digest age: ${age}s"
  echo "  preview: $(head -c 120 "$DIGEST")..."
else
  echo "✗ digest file missing — daemon may not have produced one yet"
fi

# Log size
if [ -f "$DAEMON_LOG" ]; then
  bytes=$(stat -f %z "$DAEMON_LOG" 2>/dev/null || stat -c %s "$DAEMON_LOG" 2>/dev/null)
  echo "  log: $DAEMON_LOG (${bytes} bytes)"
fi

# tmux integration?
# Display surfaces — tmux status-right + Claude Code statusLine can coexist
tmux_wired=0
statusline_wired=0

if command -v tmux >/dev/null 2>&1; then
  if grep -q claude-recap "$HOME/.tmux.conf" 2>/dev/null; then
    echo "✓ tmux status-right wired"
    tmux_wired=1
  else
    echo "○ tmux installed but status-right not wired — run /ops:recap configure"
  fi
else
  echo "○ tmux not installed"
fi

SETTINGS="$HOME/.claude/settings.json"
statusline_check_skipped=0
if ! command -v jq >/dev/null 2>&1; then
  echo "○ Claude Code statusLine — cannot check (jq not installed; install via 'brew install jq' or your distro's package manager)"
  statusline_check_skipped=1
elif [ ! -f "$SETTINGS" ]; then
  echo "○ Claude Code statusLine — no ~/.claude/settings.json found (run Claude Code at least once, or run /ops:recap configure to create one)"
  statusline_check_skipped=1
else
  sl_cmd=$(jq -r '.statusLine.command // empty' "$SETTINGS" 2>/dev/null)
  if echo "$sl_cmd" | grep -q claude-recap-digest; then
    echo "✓ Claude Code statusLine wired"
    statusline_wired=1
  else
    echo "○ Claude Code statusLine not wired — run /ops:recap configure"
  fi
fi

# Only warn "no display surface active" when we actually verified both surfaces.
# If statusline check was skipped due to missing prerequisites, statusline_wired
# being 0 is uninformative — surface that explicitly instead of misleading text.
if [ "$tmux_wired" -eq 0 ] && [ "$statusline_wired" -eq 0 ]; then
  if [ "$statusline_check_skipped" -eq 1 ]; then
    echo "  → tmux not wired; statusLine status unknown (prerequisite missing — see above). Daemon still produces /tmp/claude-recap-digest (useful for /ops:recap tail)."
  else
    echo "  → no display surface active. Daemon still produces /tmp/claude-recap-digest (useful for /ops:recap tail)."
  fi
fi

Subcommand: tail

Stream the rolling digest log:

bash
echo "─── Last 20 generated digests ───"
tail -n 20 /tmp/claude-recap-digest.log 2>/dev/null || echo "(no digest log yet)"
echo
echo "─── Daemon log (tail 30) ───"
tail -n 30 /tmp/claude-recap-daemon.log 2>/dev/null || echo "(no daemon log yet)"

Subcommand: configure

Walk the user through display-surface integration. Two surfaces are supported and they can coexist: tmux status-right and Claude Code statusLine (~/.claude/settings.json). Detect existing config first to avoid clobbering.

  1. Check tmux availability — if missing, jump to the statusLine fallback flow below:

    bash
    if ! command -v tmux >/dev/null 2>&1; then
      echo "tmux not installed — offering Claude Code statusLine fallback."
      # → see "statusLine fallback" section below
    fi
  2. Check current ~/.tmux.conf:

    bash
    TMUX_CONF="$HOME/.tmux.conf"
    tmux_already_wired=0
    if grep -q claude-recap "$TMUX_CONF" 2>/dev/null; then
      echo "✓ tmux status-right already configured."
      tmux_already_wired=1
    fi
    existing=$(grep -E '^\s*set\s+-g\s+status-right' "$TMUX_CONF" 2>/dev/null | head -1)

    Steps 3–5 only apply when tmux is not already wired (tmux_already_wired=0):

  3. If existing is non-empty, AskUserQuestion:

    • Question: "Existing tmux status-right detected: $existing. Append recap marquee?"
    • Options: [Append (keep existing)] [Replace with recap-only] [Skip]
  4. On append: insert the marquee snippet ahead of the existing setting so both render. On replace: comment out the old line and add the new one.

    Snippet to append — expand the plugin path when writing ~/.tmux.conf (tmux #() does not expand env vars):

    bash
    PLUGIN_ROOT="${CLAUDE_PLUGIN_ROOT:-$(ls -d "$HOME/.claude/plugins/cache/ops-marketplace/ops"/*/ 2>/dev/null | sort -V | tail -1)}"
    cat >> "$TMUX_CONF" <<TMUX

claude-ops recap marquee

set -g status-right '#('"${PLUGIN_ROOT}"'/scripts/recap/marquee.sh) #[fg=#a6e3a1]%H:%M ' set -g status-interval 2 TMUX

  1. Reload if tmux is running:
bash
tmux info >/dev/null 2>&1 && tmux source-file "$TMUX_CONF" 2>/dev/null

configure — statusLine fallback (no tmux, or in addition to tmux)

When tmux is missing — or the user wants a second surface — wire the recap into Claude Code's own status bar via ~/.claude/settings.json.

  1. AskUserQuestion (Rule 1 — exactly 4 options):

    Wire recap into Claude Code statusLine (~/.claude/settings.json)?
      [Add to Claude Code statusLine]  [Show me the JSON, I'll add manually]  [Skip]  [Help]
  2. On Add to Claude Code statusLine, detect existing entry:

    bash
    SETTINGS="$HOME/.claude/settings.json"
    mkdir -p "$(dirname "$SETTINGS")"
    [ -f "$SETTINGS" ] || echo '{}' > "$SETTINGS"
    existing=$(jq -r '.statusLine // empty' "$SETTINGS")
  3. If existing is non-empty, ask (Rule 1 — 3 options):

    Existing statusLine detected. How to handle?
      [Replace with recap]  [Append after current (chain commands)]  [Skip]
  4. Merge with jq so other settings are preserved:

    bash
    TMP=$(mktemp)
    jq '.statusLine = {
      "type": "command",
      "command": "cat /tmp/claude-recap-digest 2>/dev/null | head -c 80",
      "refreshInterval": 30
    }' "$SETTINGS" > "$TMP" && mv "$TMP" "$SETTINGS"

    For the append branch, chain commands so both render side-by-side:

    bash
    prev_cmd=$(jq -r '.statusLine.command // ""' "$SETTINGS")
    if [ -n "$prev_cmd" ]; then
      new_cmd="${prev_cmd}; cat /tmp/claude-recap-digest 2>/dev/null | head -c 80"
    else
      new_cmd="cat /tmp/claude-recap-digest 2>/dev/null | head -c 80"
    fi
    TMP=$(mktemp)
    jq --arg cmd "$new_cmd" '.statusLine.command = $cmd' "$SETTINGS" > "$TMP" && mv "$TMP" "$SETTINGS"
  5. Print: ✓ Claude Code statusLine wired — restart Claude Code (or open a new session) to activate.

The JSON snippet (for the manual / Show me the JSON path):

json
{
  "statusLine": {
    "type": "command",
    "command": "cat /tmp/claude-recap-digest 2>/dev/null | head -c 80",
    "refreshInterval": 30
  }
}

Subcommand: restart

Reload the launchd agent without reconfiguring:

bash
LABEL=com.claude-ops.recap-daemon
PLIST=~/Library/LaunchAgents/com.claude-ops.recap-daemon.plist
launchctl bootout "gui/$(id -u)/${LABEL}" 2>/dev/null || true
rm -rf /tmp/claude-recap-daemon.lock /tmp/claude-recap-daemon.pid
launchctl bootstrap "gui/$(id -u)" "$PLIST" && echo "✓ daemon restarted"

Linux / systemd alternative

The launchd plist is macOS-only. Linux users can wrap scripts/recap/daemon.sh in a systemd --user unit. Minimal example (~/.config/systemd/user/claude-recap-daemon.service):

ini
[Unit]
Description=Claude Ops Recap Daemon
After=default.target

[Service]
Type=simple
ExecStart=/bin/sh %h/.claude/plugins/cache/ops-marketplace/ops/CURRENT/scripts/recap/daemon.sh
Restart=always
RestartSec=30

[Install]
WantedBy=default.target

Then systemctl --user daemon-reload && systemctl --user enable --now claude-recap-daemon. The CLI subcommands above (status / tail / restart) won't auto-detect systemd — use systemctl --user status claude-recap-daemon directly.

Invocation

/ops:recap (alias: shows status). Subcommand parsing:

bash
sub="${1:-status}"
case "$sub" in
  status|tail|configure|restart) ;;
  *) echo "Unknown subcommand: $sub. Try: status | tail | configure | restart"; exit 1 ;;
esac

Frequently asked questions

What does the Ops Recap AI skill do?

OPS on-demand: This skill should be used when the user asks to "recap daemon", "tmux marquee", or…

Why use Ops Recap on TypingMind?

Because you install it once and use it with any model. Ops Recap is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Ops Recap in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/Lifecycle-Innovations-Limited/claude-ops/tree/main/claude-ops/skills/ops-recap. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Ops Recap?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Ops Recap?

As many as you like. As long as a model supports skills, you can use Ops Recap with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Ops Recap AI skill free?

Yes. It is published on GitHub by Lifecycle-Innovations-Limited under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇