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 skills/bikeread/promethos/debug-agent-failuresnpx skills add bikeread/promethos --skill debug-agent-failuresgit clone --depth 1 https://github.com/bikeread/promethosWhat 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 | $0.00031 | $0.00822 |
| Opus 5 | $0.00015 | $0.00411 |
| Sonnet 5 | $0.00006 | $0.00164 |
| Haiku 4.5 | $0.00003 | $0.00082 |
Grade A, and why
debug-agent-failures 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 3d 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.
How it starts
The opening of the file, as written. The whole thing — 95 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Goal
Move from a vague symptom to a defensible root-cause hypothesis and a small next fix.
Inputs
- Failure symptom or transcript
- Relevant logs, prompts, traces, or outputs
- Current implementation context
Non-Goals
- Jumping straight to fixes because they sound plausible
- Rewriting large parts of the system without isolating the failure mode
Workflow
Trigger signals (for proactive recognition)
- The same tool call fails more than once
- The agent loops back to a step it already tried
- The user says "又出错了" or "it broke again"
- The user says "it keeps looping", "it got worse after the change", or "the same failure keeps coming back"
- Output quality degrades after a change
- A hallucination appears in tool-calling arguments
1. Reproduce and classify the symptom
State what went wrong, under what conditions it happens, and whether it looks like a planning, context, tool, permission, memory, or orchestration failure. Success criteria: The bug is framed as a reproducible failure class rather than a vague impression.
2. Collect the smallest decisive evidence
Inspect the transcript, prompts, tool calls, logs, state transitions, and input artifacts needed to narrow the search space without drowning in noise. If the current workspace does not actually contain the target system's code, logs, traces, config, or transcripts, stop and ask for the real evidence source before widening the search. Do not substitute broad searches across unrelated repositories for the missing local evidence. Success criteria: The likely causes are constrained by actual evidence.
3. Form competing root-cause hypotheses
List the smallest plausible explanations and note what evidence would confirm or disprove each one. Success criteria: Debugging has explicit hypotheses rather than a single unchallenged guess.
4. Test the leading hypothesis with the smallest intervention
Choose the least invasive change, probe, or experiment that can separate the best explanation from the rest. Success criteria: There is a clear next action tied to the strongest hypothesis.
What ships with it
1 file 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.
- 3d ago First seen · 95 lines · 31 tokens per session scan A 4d9a817b899e
debug-agent-failures is a skill published in the GitHub repository bikeread/promethos (33 stars, last pushed 5mo ago), licensed MIT. It adds 31 tokens to every session and 822 once invoked, about $0.0002 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 2026-08-30.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
auto-perf-optimize
Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.
chat-perf
Run chat perf benchmarks and memory leak checks against the local dev build or any published VS Code version. Use when investigating chat rendering regressions, validating perf-sensitive changes to chat UI, or checking for memory leaks in the chat response pipeline.
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…