ember-trace skill
✓ VerifiedPerforms runtime forensics on system artifacts like cpuprofiles, traces, and heap snapshots to provide cited diagnoses of performance issues.
- GitHub stars
- 113
- Files
- 3
- Last commit
- Nov 1, 2026
- License
- MIT
- Allowed tools
- —
What does the ember-trace skill do?
ember-trace performs runtime forensics on system artifacts like cpuprofiles, traces, and heap snapshots to provide cited diagnoses of performance issues.
Summary generated from the skill's SKILL.md.
About the ember-trace skill
Performing runtime forensics on system artifacts to find the source of performance issues is what ember-trace enables. Developers use this skill to investigate memory leaks, idle CPU spins, visual glitches, or system hangs by analyzing the evidence left behind by a running system. It is designed for situations where you have a captured cpuprofile, a .json.gz trace, a spindump, or a heap snapshot, or when you are observing a live process exhibiting symptoms.
An agent should load this skill when it needs to move beyond theorizing from source code to reading actual system records. The skill produces a cited diagnosis that maps findings to specific files, symbols, and line numbers. It does not provide code fixes, but instead identifies the exact line that allocates, schedules, or blocks. To protect the context window, large artifacts are parsed away from the main thread by dispatching the work to team-avatar:zuko. Once a diagnosis is complete, the results can be handed to healing-water for defects or breath-of-fire for slowness.
What an agent can do with it
- 1Diagnose a memory leakAn AI client can analyze a heap snapshot to find the specific line of code causing memory exhaustion.
- 2Investigate CPU performance issuesA developer can provide a cpuprofile to identify why a system is experiencing an idle CPU spin.
Written from the skill's SKILL.md. Every use case is checked against a quote from it.
How to install the ember-trace skill
Each command copies the skill folder into the client's skills directory. Restart the client or start a new session to load it.Install the ember-trace skill in Plugin
/plugin marketplace add tmcfarlane/team-avatar /plugin install team-avatar@team-avatar
Install the ember-trace skill in Claude Code
~/.claude/skills/ember-trace/SKILL.md
$ npx degit tmcfarlane/team-avatar/plugins/team-avatar/skills/ember-trace ~/.claude/skills/ember-traceInstall the ember-trace skill in Claude Code (project)
.claude/skills/ember-trace/SKILL.md
$ npx degit tmcfarlane/team-avatar/plugins/team-avatar/skills/ember-trace .claude/skills/ember-traceInstall the ember-trace skill in Codex
~/.agents/skills/ember-trace/SKILL.md
$ npx degit tmcfarlane/team-avatar/plugins/team-avatar/skills/ember-trace ~/.agents/skills/ember-traceInstall the ember-trace skill in Gemini CLI
~/.gemini/skills/ember-trace/SKILL.md
$ npx degit tmcfarlane/team-avatar/plugins/team-avatar/skills/ember-trace ~/.gemini/skills/ember-traceInstall the ember-trace skill in Claude.ai
npx degit tmcfarlane/team-avatar/plugins/team-avatar/skills/ember-trace ember-trace && zip -r ember-trace.zip ember-trace # claude.ai → Customize → Skills → + → Create skill → Upload a skill → ember-trace.zip
Automated checks
Ran Nov 14, 2026Full report →Passed our automated checks. That is not a security audit: read the scripts before you let an agent run them. How checks work
- Source repository linked and reachabletmcfarlane/team-avatar/plugins/team-avatar/skills/ember-trace is public and reachablegithub.com/tmcfarlane/team-avatar/tree/HEAD/plugins/team-ava ↗Pass
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit 2026-11-01github.com/tmcfarlane/team-avatar/commits ↗Pass
- No download-and-run or credential theft in scriptsNo scripts in the skill folderPass
- No hidden instructions in SKILL.mdNo override phrases, safeguard switches, credential requests or hidden charactersgithub.com/tmcfarlane/team-avatar/blob/HEAD/plugins/team-ava ↗Pass
- Pre-approved tools are scopedNo pre-approved tools; the agent asks before using onegithub.com/tmcfarlane/team-avatar/blob/HEAD/plugins/team-ava ↗Pass
- SKILL.md follows the Agent Skills formatValid name and description; says when to use it; 34 linesgithub.com/tmcfarlane/team-avatar/blob/HEAD/plugins/team-ava ↗Pass
SKILL.md
View on GitHubZuko's ember trace
When Zuko tracked the Avatar, he did not guess where the fire started. He read the scorch marks and followed the embers back to their source. Forensics is the same. You do not theorize from source code. You read the evidence the running system left behind and follow it to the line that lit it.
You own the diagnosis. The deliverable is a cited diagnosis, not a fix. Fix only when asked. Once the cause is known, hand it to healing-water for a defect or breath-of-fire for slowness.
On Codex, read avatar-mode/references/codex.md for tool equivalents.
Pick the mode
| You have | Mode | Steps |
|---|---|---|
| A live process showing the symptom now | Live | references/live.md |
A captured artifact (cpuprofile, .json.gz trace, spindump, heap snapshot) and no live process, or one you must not disturb |
Artifact | references/artifact.md |
If you have both, start from the artifact and use the live process to prove the mechanism.
Rules for both modes
- A real artifact, not a guess. The evidence is what the system recorded, not what the source suggests.
- Parse large artifacts away from the main thread. When the top-level session runs this skill, dispatch the parsing to
team-avatar:zukoon Zuko's model from the avatar-setup sheet, with an effort level that fits the job, and keep only the reduced finding in the main thread (Guard the Context Window, Fire infour-elements). When you are already a dispatched character, reduce it yourself with a script and read only the script's output. - Map every finding to source: file, symbol, and the line that allocates, schedules, or blocks. A frame with no source location is not yet a diagnosis.
- Label confidence. A mechanism you proved is a cause. One the evidence only supports is the strongest hypothesis, and you say so.
- No change set ships from this skill, so there is no throughput or PR step.
Reply
The signal or artifact and its format, the reduced finding, how you proved the mechanism (or that it is the strongest hypothesis), the source location, and the artifact paths. No fix unless asked.
Files in this skill
3ember-trace skill questions
How do I install the ember-trace skill in Claude Code?
Run npx degit tmcfarlane/team-avatar/plugins/team-avatar/skills/ember-trace ~/.claude/skills/ember-trace. Claude Code loads skills from ~/.claude/skills (all projects) or .claude/skills (one project) and uses a skill when its description matches the task.
Which clients can use the ember-trace skill?
Any client that reads the Agent Skills format: Claude Code, Codex, Gemini CLI and claude.ai (upload as a ZIP). Scripts in the skill run only where the client allows code execution.
What is the final deliverable of this skill?
The skill provides a cited diagnosis of the issue rather than a code fix.
What modes of operation are available?
You can use Live mode for a live process showing symptoms or Artifact mode for captured files like cpuprofiles or heap snapshots.
What information must be included in the reply?
The reply should include the signal or artifact format, the reduced finding, the proof of the mechanism, the source location, and artifact paths.