Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/generous-corp/pulp/tracegit clone --depth 1 https://github.com/Generous-Corp/pulpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/generous-corp/pulp/trace)<a href="https://agentmods.dev/commands/generous-corp/pulp/trace"><img src="https://agentmods.dev/badge/commands/generous-corp/pulp/trace.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00029 | $0.01156 |
| Opus 5 | $0.00015 | $0.00578 |
| Sonnet 5 | $0.00006 | $0.00231 |
| Haiku 4.5 | $0.00003 | $0.00116 |
Grade A, and why
trace scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 107 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Answer "why is this slow?" from Perfetto evidence. Offline .pftrace analysis
is usable today. Live lifecycle control is broker-authorized and fail-closed.
Tracing is a dev-only tool. Never ship a plugin with PULP_TRACING enabled.
Five paths (pick by what you have)
| You have | Path | Tool |
|---|---|---|
| Authorized control session + want a DSP flamegraph | Experimental live trace | Start/stop through canonical control |
| An A3 GPU startup trace + bounded question | Named A2T analysis | /trace gpu-startup|gpu-health|gpu-probe --trace FILE.pftrace |
A .pftrace + a question |
Query | pulp trace query "<sql>" --trace FILE.pftrace |
| Want to hand an agent / human the raw file | Return the path | pulp trace stop prints it; open in ui.perfetto.dev |
| A UI hitch to correlate | Frame trace + motion join | trace render,layout while a motion trace runs; query on shared trace_id |
Experimental live path
# The broker must authorize a trusted control session.
# 2. Start a session, reproduce the slow thing, then stop. Use the same exact
# broker-owned instance ID for both lifecycle calls when selecting explicitly.
pulp trace start --instance INSTANCE_ID --categories dsp,render
# ... trigger the suspect interaction / open the editor ...
pulp trace stop --instance INSTANCE_ID
# → /tmp/pulp-<ts>.pftrace
# 3. Query the flushed file. For a narrated answer, load trace-analysis and
# use its trace-sql queries over this same file.
pulp trace query "SELECT name, dur FROM slice ORDER BY dur DESC LIMIT 20" \
--trace /tmp/pulp-trace.pftrace
Named A2T GPU analysis
Use the slash command with one exact bounded question; it delegates to the same typed installed analysis surface used by the CLI/MCP parity receipt:
/trace gpu-startup --trace /tmp/pulp.pftrace
/trace gpu-health --trace /tmp/pulp.pftrace
/trace gpu-probe --trace /tmp/pulp.pftrace
The underlying reproducible commands are:
pulp trace gpu-startup --trace /tmp/pulp.pftrace --json
pulp trace gpu-health --trace /tmp/pulp.pftrace --json
pulp trace gpu-probe --trace /tmp/pulp.pftrace --json
pulp trace open /tmp/pulp.pftrace
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- today First seen · 107 lines · 29 tokens per session scan A b57e63a2a52e
trace is a command published in the GitHub repository Generous-Corp/pulp (16 stars, last pushed today), licensed MIT. It adds 29 tokens to every session and 1,156 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-04.
Other commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.