feat: show teammate model in the agent panel by GordonBeeming · Pull Request #14 · GordonBeeming/claude-statusline

GordonBeeming · 2026-07-08T09:47:34Z

What

Background teammates show up as rows in Claude Code's agent panel, but the default row (name · description · token count) never says which model a teammate is running: a Sonnet teammate looks the same as an Opus one.

This adds subagent-statusline.sh, a subagentStatusLine command that rewrites each running teammate's row:

↳ 🐝 pr-pilot · 🔮 Sonnet 5 · ↓ 81.8k tok
↳ 🐝 explorer · 🍃 Haiku 4.5 · ↓ 12.1k tok

↳ 🐝 marks a spawned worker nested under the main session.
The model comes from the teammate's own transcript, since the panel payload has no model field. It fills in once the teammate's first response lands.
Only actively-running teammates show; finished and idle rows are hidden, so the panel goes quiet once work is done.

Glyphs: 🦉 Opus · 🔮 Sonnet · 🍃 Haiku · 📖 Fable/Mythos. A model outside these families keeps its row without a glyph.

Also

statusline.sh's once-a-day auto-update now refreshes both scripts, so the subagent script never goes stale after install.
install.sh installs both scripts and prints the subagentStatusLine snippet.
README documents the feature.

Testing

Verified live through the configured command against real captured payloads:

tok=0     → ↳ 🐝 <name> · ↓ 0 tok                    (just spawned, no model yet)
tok=20504 → ↳ 🐝 <name> · 🦉 Opus 4.8 · ↓ 20.5k tok

Full model matrix (Opus, Sonnet, Haiku, Fable, plus legacy Haiku 3.5) and the edge cases all pass: unknown family keeps the row without a glyph, a completed teammate's row is hidden, and sub-1k token counts render raw. All three scripts pass bash -n.

Note: existing sessions load subagentStatusLine at startup, so restart or open a new session to pick it up.

🤖 Generated with Claude Code

https://claude.ai/code/session_01EHAoGEYEfzyhPRHSQ1XkDZ

Add subagent-statusline.sh, a subagentStatusLine command that rewrites each running teammate's row in the agent panel as "↳ 🐝 <name> · <glyph> <model> · ↓ <tok> tok". The model is read from the teammate's own transcript (the panel payload has no model field) and mapped to a per-family glyph (Opus/Sonnet/Haiku/Fable). Only actively-running teammates show; finished and idle rows are hidden. statusline.sh's daily auto-update now refreshes both scripts, and install.sh installs both and prints the subagentStatusLine snippet. Co-authored-by: Claude <noreply@anthropic.com> Co-authored-by: GitButler <gitbutler@gitbutler.com>

The Status Line Example now shows the four-line main block and the agent-panel teammate rows together, so the difference is visible at a glance. Drop the duplicate example from the Subagent rows section and link back to it. Co-authored-by: Claude <noreply@anthropic.com> Co-authored-by: GitButler <gitbutler@gitbutler.com>

gemini-code-assist

Code Review

This pull request introduces a new subagent-statusline.sh script to display active subagent/teammate models and token usage in the Claude Code agent panel, along with corresponding installer and auto-update integrations. The review feedback focuses on performance optimizations and robust error handling under set -e. Key recommendations include optimizing transcript scanning by only reading the tail of the file, replacing external awk calls with pure Bash arithmetic for token formatting, declaring loop variables as local, and safely capturing command substitution exit codes to prevent premature script termination.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

gemini-code-assist · 2026-07-08T09:50:53Z

+  grep -oE '"model":"claude[^"]*"' "$file" 2>/dev/null \
+    | tail -1 \
+    | sed -E 's/.*"model":"//; s/"$//' || true


Since this script runs on every panel refresh tick, scanning the entire teammate transcript file with grep can cause noticeable terminal lag as the transcript grows large. Since the model ID is repeated in recent messages, optimize this by only scanning the last 50 lines of the file using tail.

Suggested change

grep -oE '"model":"claude[^"]*"' "$file" 2>/dev/null \

| tail -1 \

| sed -E 's/.*"model":"//; s/"$//' || true

tail -n 50 "$file" 2>/dev/null \

| grep -oE '"model":"claude[^"]*"' \

| tail -1 \

| sed -E 's/.*"model":"//; s/"$//' || true

References

In shell scripts running with 'set -e', guard command substitutions or pipelines that may fail with '|| true' to prevent script termination, provided that an empty result is safely handled by fallback logic.

gemini-code-assist · 2026-07-08T09:50:53Z

+transcript_path=$(printf '%s' "$stdin_data" | jq -r '.transcript_path // ""' 2>/dev/null || echo "")
+columns=$(printf '%s' "$stdin_data" | jq -r '.columns // 80' 2>/dev/null || echo 80)
+[[ "$columns" =~ ^[0-9]+$ ]] || columns=80


Capture command substitution exit codes using the || status=$? pattern to prevent unintended script termination on command failure under set -e.

Suggested change

transcript_path=$(printf '%s' "$stdin_data" | jq -r '.transcript_path // ""' 2>/dev/null || echo "")

columns=$(printf '%s' "$stdin_data" | jq -r '.columns // 80' 2>/dev/null || echo 80)

[[ "$columns" =~ ^[0-9]+$ ]] || columns=80

status=0

transcript_path=$(printf '%s' "$stdin_data" | jq -r '.transcript_path // ""' 2>/dev/null) || status=$?

columns=$(printf '%s' "$stdin_data" | jq -r '.columns // 80' 2>/dev/null) || status=$?

[[ "$columns" =~ ^[0-9]+$ ]] || columns=80

References

When using 'set -e' in shell scripts, capture command substitution exit codes using the 'var=$(command) || status=$?' pattern to prevent unintended script termination on command failure.

gemini-code-assist · 2026-07-08T09:50:53Z

+  local first rest
+  first=$(printf '%s' "${family:0:1}" | tr '[:lower:]' '[:upper:]')


Under set -euo pipefail, command substitutions can cause unintended script termination if any command in the pipeline fails. Capture the exit code using the || status=$? pattern to prevent this.

Suggested change

local first rest

first=$(printf '%s' "${family:0:1}" | tr '[:lower:]' '[:upper:]')

local first rest status=0

first=$(printf '%s' "${family:0:1}" | tr '[:lower:]' '[:upper:]') || status=$?

References

When using 'set -e' in shell scripts, capture command substitution exit codes using the 'var=$(command) || status=$?' pattern to prevent unintended script termination on command failure.

gemini-code-assist · 2026-07-08T09:50:53Z

+fmt_tokens() {
+  local n="$1"
+  [[ "$n" =~ ^[0-9]+$ ]] || n=0
+  if (( n >= 1000 )); then
+    awk -v n="$n" 'BEGIN { printf "%.1fk", n/1000 }'
+  else
+    printf '%s' "$n"
+  fi
+}


Spawning an external awk process on every panel refresh tick inside a loop can be a performance bottleneck. Implement token formatting and rounding entirely in pure Bash, which is significantly faster and avoids process forks.

Suggested change

fmt_tokens() {

local n="$1"

[[ "$n" =~ ^[0-9]+$ ]] || n=0

if (( n >= 1000 )); then

awk -v n="$n" 'BEGIN { printf "%.1fk", n/1000 }'

else

printf '%s' "$n"

fi

}

fmt_tokens() {

local n="$1"

[[ "$n" =~ ^[0-9]+$ ]] || n=0

if (( n >= 1000 )); then

local rounded=$(( (n + 50) / 100 ))

printf '%d.%dk' $(( rounded / 10 )) $(( rounded % 10 ))

else

printf '%s' "$n"

fi

}

gemini-code-assist · 2026-07-08T09:50:53Z

+tasks_tsv=$(printf '%s' "$stdin_data" | jq -r '
+  .tasks // []
+  | .[]
+  | [ .id, (.status // ""), (.label // .description // ""), ((.tokenCount // 0) | tostring) ]
+  | @tsv
+' 2>/dev/null || true)


Capture command substitution exit codes using the || status=$? pattern to prevent unintended script termination on command failure under set -e.

Suggested change

tasks_tsv=$(printf '%s' "$stdin_data" | jq -r '

.tasks // []

| .[]

| [ .id, (.status // ""), (.label // .description // ""), ((.tokenCount // 0) | tostring) ]

| @tsv

' 2>/dev/null || true)

status=0

tasks_tsv=$(printf '%s' "$stdin_data" | jq -r '

.tasks // []

| .[]

| [ .id, (.status // ""), (.label // .description // ""), ((.tokenCount // 0) | tostring) ]

| @tsv

' 2>/dev/null) || status=$?

References

When using 'set -e' in shell scripts, capture command substitution exit codes using the 'var=$(command) || status=$?' pattern to prevent unintended script termination on command failure.

gemini-code-assist · 2026-07-08T09:50:53Z

+  raw_model=$(model_of "$id")
+  if [[ -n "$raw_model" ]]; then
+    glyph=$(model_glyph "$raw_model")
+    if pretty=$(model_pretty "$raw_model"); then
+      if [[ -n "$glyph" ]]; then
+        content="${content} · ${glyph} ${pretty}"
+      else
+        content="${content} · ${pretty}"
+      fi
+    fi
+  fi
+
+  content="${content} · ↓ $(fmt_tokens "$tok") tok"


Declare loop variables as local to prevent scope leakage, and capture command substitution exit codes using the || status=$? pattern to adhere to the general rules and prevent unintended script termination under set -e.

Suggested change

raw_model=$(model_of "$id")

if [[ -n "$raw_model" ]]; then

glyph=$(model_glyph "$raw_model")

if pretty=$(model_pretty "$raw_model"); then

if [[ -n "$glyph" ]]; then

content="${content} · ${glyph} ${pretty}"

else

content="${content} · ${pretty}"

fi

fi

fi

content="${content} · ↓ $(fmt_tokens "$tok") tok"

local raw_model="" glyph="" pretty="" formatted_tok="" status=0

raw_model=$(model_of "$id") || status=$?

if [[ -n "$raw_model" ]]; then

glyph=$(model_glyph "$raw_model") || status=$?

pretty=$(model_pretty "$raw_model") || status=$?

if [[ $status -eq 0 && -n "$pretty" ]]; then

if [[ -n "$glyph" ]]; then

content="${content} · ${glyph} ${pretty}"

else

content="${content} · ${pretty}"

fi

fi

fi

formatted_tok=$(fmt_tokens "$tok") || status=$?

content="${content} · ↓ ${formatted_tok} tok"

References

When using 'set -e' in shell scripts, capture command substitution exit codes using the 'var=$(command) || status=$?' pattern to prevent unintended script termination on command failure.

Copilot

Pull request overview

Adds a dedicated subagentStatusLine command to improve Claude Code’s agent panel by showing each running teammate’s model (with optional glyphs), and updates install/auto-update flows so both statusline scripts stay current.

Changes:

Introduces subagent-statusline.sh to render per-subagent rows including model family/version and compact token counts.
Extends statusline.sh daily auto-update to refresh both statusline.sh and subagent-statusline.sh.
Updates install.sh and README.md to install/document the optional subagentStatusLine configuration.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 3 comments.

File	Description
subagent-statusline.sh	New script that parses tasks + subagent transcripts to render model-aware agent panel rows.
statusline.sh	Auto-update now refreshes both the main and subagent statusline scripts.
install.sh	Installer now installs both scripts and prints config snippets for both.
README.md	Documents the new subagent rows feature and updated install/config instructions.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

+  grep -oE '"model":"claude[^"]*"' "$file" 2>/dev/null \
+    | tail -1 \
+    | sed -E 's/.*"model":"//; s/"$//' || true


+  if curl -sSL "${BASE_URL}/${name}" -o "$tmp"; then
+    cp "$tmp" "${INSTALL_DIR}/${name}"
+    chmod +x "${INSTALL_DIR}/${name}"
+    echo "  Installed successfully."


+model_of() {
+  local id="$1" file="${subagents_dir}/agent-${id}.jsonl"
+  [[ -f "$file" ]] || return 0
+  grep -oE '"model":"claude[^"]*"' "$file" 2>/dev/null \
+    | tail -1 \
+    | sed -E 's/.*"model":"//; s/"$//' || true


Copilot

Pull request overview

Copilot reviewed 4 out of 4 changed files in this pull request and generated 3 comments.

+columns=$(printf '%s' "$stdin_data" | jq -r '.columns // 80' 2>/dev/null || echo 80)
+[[ "$columns" =~ ^[0-9]+$ ]] || columns=80


+  grep -oE '"model":"claude[^"]*"' "$file" 2>/dev/null \
+    | tail -1 \
+    | sed -E 's/.*"model":"//; s/"$//' || true


+  echo "Installing ${name} to ${INSTALL_DIR}..."
+  local tmp
+  tmp=$(mktemp)
+  if curl -sSL "${BASE_URL}/${name}" -o "$tmp"; then


chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 42250e150a

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

chatgpt-codex-connector · 2026-07-08T09:55:53Z

+# Last assistant model recorded in a teammate's transcript. Empty if the file is
+# missing or no assistant turn has landed yet (a just-spawned agent).
+model_of() {
+  local id="$1" file="${subagents_dir}/agent-${id}.jsonl"


Resolve teammate sessions before reading the model

When the row represents an Agent Teams teammate, .tasks[].id is the panel/task id while the teammate is a separate Claude session whose session id is recorded in the team config; its transcript is not <lead-session>/subagents/agent-$id.jsonl. With this hard-coded path [[ -f "$file" ]] returns false, so raw_model stays empty and the model never appears for the teammate rows this feature is meant to improve. Please map teammate rows to their session transcript and only use the subagents path for classic subagents.

Useful? React with 👍 / 👎.

GordonBeeming marked this pull request as ready for review July 8, 2026 09:48

Copilot AI review requested due to automatic review settings July 8, 2026 09:48

Copilot started reviewing on behalf of GordonBeeming July 8, 2026 09:48 View session

gemini-code-assist Bot reviewed Jul 8, 2026

View reviewed changes

Copilot AI reviewed Jul 8, 2026

View reviewed changes

Copilot AI review requested due to automatic review settings July 8, 2026 09:51

Copilot started reviewing on behalf of GordonBeeming July 8, 2026 09:51 View session

Copilot AI reviewed Jul 8, 2026

View reviewed changes

chatgpt-codex-connector Bot reviewed Jul 8, 2026

View reviewed changes

GordonBeeming merged commit 3f7eafb into main Jul 8, 2026
1 check passed

GordonBeeming deleted the gb/subagent-model-row branch July 8, 2026 10:23

Navigation Menu

Repository navigation

feat: show teammate model in the agent panel - #14

feat: show teammate model in the agent panel#14
GordonBeeming merged 2 commits into
mainfrom
gb/subagent-model-row

GordonBeeming commented Jul 8, 2026

Uh oh!

gemini-code-assist Bot left a comment

Uh oh!

gemini-code-assist Bot Jul 8, 2026

Uh oh!

gemini-code-assist Bot Jul 8, 2026

Uh oh!

gemini-code-assist Bot Jul 8, 2026

Uh oh!

gemini-code-assist Bot Jul 8, 2026

Uh oh!

gemini-code-assist Bot Jul 8, 2026

Uh oh!

gemini-code-assist Bot Jul 8, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI left a comment

Uh oh!

chatgpt-codex-connector Bot left a comment

Uh oh!

chatgpt-codex-connector Bot Jul 8, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

		local first rest
		first=$(printf '%s' "${family:0:1}" \| tr '[:lower:]' '[:upper:]')

		columns=$(printf '%s' "$stdin_data" \| jq -r '.columns // 80' 2>/dev/null \|\| echo 80)
		[[ "$columns" =~ ^[0-9]+$ ]] \|\| columns=80

Conversation

GordonBeeming commented Jul 8, 2026

What

Also

Testing

Uh oh!

gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

gemini-code-assist Bot Jul 8, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jul 8, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jul 8, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jul 8, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jul 8, 2026

Choose a reason for hiding this comment

Uh oh!

gemini-code-assist Bot Jul 8, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

chatgpt-codex-connector Bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

chatgpt-codex-connector Bot Jul 8, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants