perfetto
Skill9 starsChanged 4 days agoMITView on GitHub
Use when the user wants to view, inspect, or share Perfetto/Chrome trace files -- triggered by words like "trace", "perf", "perfetto", "profiling", "trace viewer", or when dealing with trace JSON files from training runs.
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 14
All URLs use <base_url> where <base_url> is the CharlieBot URL resolved from HOST MEMORY (look for the CharlieBot URL entry).
- Distinctive: line 31
If a directory contains multiple time windows (or multiple steps) for the same rank(s), generate
- Very specific: line 69
Before using ?dir=, confirm the directory holds at most one trace per rank for a single time window.
- Distinctive: line 71
Only JSON traces can be merged.
≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Perfetto Trace Viewer
Generate links to the CharlieBot built-in Perfetto trace viewer, which can merge cross-rank traces of the same time window into a single view.
Endpoints
All URLs use <base_url> where <base_url> is the CharlieBot URL resolved from HOST MEMORY (look
for the CharlieBot URL entry).
Merge cross-rank traces from a directory
<base_url>/perfetto?dir=<absolute-path>Auto-discovers all *.json trace files in the directory, merges them client-side (labeling by
rank), and loads into an embedded Perfetto UI.
Use ?dir= only for cross-rank merging, that is, when every file in the directory is a
different rank's trace of the same time window/step. The merger labels tracks by rank and
assumes one trace per rank; mixing multiple time windows of the same rank into one ?dir= link will
produce a confusing/overlapping view.
If a directory contains multiple time windows (or multiple steps) for the same rank(s), generate
separate links, one per time window, each pointing at just that window's files. Use ?pattern=
to scope a ?dir= link to a single window, or list the per-rank files explicitly with repeated
?trace= params.
Optional params:
&pattern=<glob>: file pattern to match (default:*.json); use this to isolate one time window in a mixed directory&title=<text>: custom page title
View specific trace files
<base_url>/perfetto?trace=/absolute_filepath/<path1>&trace=/absolute_filepath/<path2>Pass one or more trace= params, each naming a trace file the way the file-server skill writes a
file link.
Single trace
<base_url>/perfetto?trace=/absolute_filepath/<absolute-path>Examples
| Scenario | URL |
| Cross-rank merge for one step | <base_url>/perfetto?dir=/path/to/traces/step100 |
Scope ?dir= to one time window | <base_url>/perfetto?dir=/path/to/traces&pattern=trace_rank*_step100.json |
| Only rank 0 | <base_url>/perfetto?trace=/absolute_filepath/path/to/traces/trace_rank000_step100.json |
| Two time windows of rank 0 | generate two separate links, one per window, rather than one combined ?dir= |
Rules
?dir=is for cross-rank merging only. Multiple time-window traces of the same rank belong in separate links, one URL per window, rather than bundled in one?dir=- Before using
?dir=, confirm the directory holds at most one trace per rank for a single time window; otherwise narrow it with&pattern=or switch to explicit?trace=params - Only JSON traces can be merged; a protobuf trace shows only the first file
In this repository21 other files it gives its agents.
- AGENTS.mdAGENTS.md
- charliebotSkill
- claude-versionSkill
- cli-tool-trapsSkill
- feishuSkill
- file-serverSkill
- gmailSkill
- google-docsSkill
- google-oauthSkill
- google-sheetsSkill
- improve-goalSkill
- improve-workerSkill
- language-portSkill
- linearSkill
- llm-context-guidelineSkill
- plan-approvalSkill
- pytorch-gpu-trainingSkill
- skill-managementSkill
- slackSkill
- voice-notesSkill
- writing-styleSkill
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.