Borrowing it
Nothing to install: this file belongs to OlegHQ/agentpack. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/OlegHQ/agentpack/dev/.agents/skills/analyze-proxy-log/SKILL.mdgit clone --depth 1 https://github.com/OlegHQ/agentpackWrote 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/skills/oleghq/agentpack/analyze-proxy-log)<a href="https://agentmods.dev/skills/oleghq/agentpack/analyze-proxy-log"><img src="https://agentmods.dev/badge/skills/oleghq/agentpack/analyze-proxy-log/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/oleghq/agentpack/analyze-proxy-log"><img src="https://agentmods.dev/badge/skills/oleghq/agentpack/analyze-proxy-log.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 5.1 | $0.00081 | $0.00473 |
| Opus 5.5 | $0.00032 | $0.00189 |
| Sonnet 5.5 | $0.00016 | $0.00095 |
| Haiku 4.5 | $0.00008 | $0.00047 |
Grade A, and why
analyze-proxy-log 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 2mo ago.
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.
What it actually says
Analyze Proxy Log
Use this skill to quickly diagnose agentpack --proxy claude failures from the proxy JSONL log.
Workflow
- Locate the log:
- Prefer an explicit path from the user.
- Else read
AGENTPACK_PROXY_LOG_DIR. - Else find the nearest agentpack project root and use
$AGENTPACK_HOME/projects/<project-hash>/proxy-logs/latest.json.
- Run
scripts/analyze_proxy_log.pywith the log file or log directory. - Read the summary first, then inspect the request timelines flagged as
LIKELY ISSUE. - Classify the primary failure:
stalled_request: request accepted but no terminal event.websocket_timeout: connect/read/write timeout or closed-before-terminal WebSocket.websocket_setup: 401/403/429 or setup frame error before a normal response.auth_refresh: refresh attempted and failed or repeated authorization failure.upstream_http: non-success HTTP status from Codex.translation: request or response conversion failed.downstream: Claude disconnected or response write failed.
- Report the likely root cause, the request ID, transport, model mapping, first failing event, and the exact log path.
Commands
python3 .agents/skills/analyze-proxy-log/scripts/analyze_proxy_log.py
python3 .agents/skills/analyze-proxy-log/scripts/analyze_proxy_log.py "$AGENTPACK_PROXY_LOG_DIR"
python3 .agents/skills/analyze-proxy-log/scripts/analyze_proxy_log.py /path/to/proxy-*.jsonl
Do not paste full payload snippets into the final answer unless the user explicitly asks. Logs are sanitized by default, but payload logging may be enabled with AGENTPACK_PROXY_LOG_PAYLOADS=1.
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 2mo ago First seen · 37 lines · 81 tokens per session scan A 75ac3b04b897
analyze-proxy-log is a skill published in the GitHub repository OlegHQ/agentpack (16 stars, last pushed 22d ago), licensed MIT. It adds 81 tokens to every session and 473 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-08-30.
analyze-proxy-log compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| analyze-proxy-logOlegHQ/agentpack | Skill | 16 | 22d ago | MIT | A | 81 | Codex |
| systematic-debuggingdamianvtran/local-operator | Skill | 229 | yesterday | MIT | A | 45 | Claude Code, Codex |
| regexbug-ops/zeph | Skill | 66 | 3d ago | MIT | A | 85 | Claude Code, Codex |
| code-analysisbug-ops/zeph | Skill | 66 | 3d ago | MIT | A | 105 | Claude Code, Codex |
| networkbug-ops/zeph | Skill | 66 | 3d ago | MIT | A | 83 | Claude Code, Codex |
| sephera-change-safetyReim-developer/Sephera | Skill | 77 | 7d ago | GPL-3.0 | A | 80 | Claude Code, Codex |
| tracelet-instrumentjnMetaCode/skillet | Skill | 2 | 1mo ago | MIT | A | 51 | Claude Code, Codex |