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 skills add partikularwaters/Banka --skill dredgegit clone --depth 1 https://github.com/partikularwaters/BankaWrote 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/partikularwaters/banka/dredge)<a href="https://agentmods.dev/skills/partikularwaters/banka/dredge"><img src="https://agentmods.dev/badge/skills/partikularwaters/banka/dredge/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/partikularwaters/banka/dredge"><img src="https://agentmods.dev/badge/skills/partikularwaters/banka/dredge.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00059 | $0.01212 |
| Opus 5.5 | $0.00024 | $0.00485 |
| Sonnet 5.5 | $0.00012 | $0.00242 |
| Haiku 4.5 | $0.00006 | $0.00121 |
Grade A, and why
dredge 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 1mo 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 — 81 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Not every problem is a bug. Not every bug needs debugging. Diagnose the failure first, then prescribe the response — two separate steps, never swapped.
Context Contract
Required: the user's failure description · once diagnosed, the implicated file/function.
Conditional: an already resolved Banka classification, only in the Hard Reset path, to determine how to phrase the next-session handoff · wider codebase context, only if a first fix fails and re-diagnosis is needed.
Excluded by default: the codebase beyond what the diagnosis implicates · no upfront state-resolution matrix — dredge doesn't run one.
Outputs: a diagnosed failure mode with reasoning, plus a mode-specific
proposal held for confirmation — or, when reproduction itself can't be
confirmed, an [OPEN — needs verification] flag and a stop instead of a
proposal.
Write authority: none.
Step 1 — Describe What Went Wrong
Describe what is wrong. Be specific:
- What did you expect to happen?
- What happened instead?
- How many times have you tried to fix it already?
The number of fix attempts matters — it distinguishes a fresh problem from a session that's already gone wrong.
Step 2 — Identify the Failure Mode
Mode 0 — Cannot reproduce. The report describes a real interaction, but a genuine attempt to trigger it — following the same steps — does not produce the described behavior. → Not yet confirmed as any of the three modes below; diagnose the gap itself first. Response: Reproduction diagnosis (Step 3D).
Mode 1 — A specific thing is broken. Isolated, rest of the project works, first/second attempt, error is clear and specific. → Normal bug, root cause is findable. Response: Targeted fix (Step 3A).
Mode 2 — The session has gone wrong. Multiple fix attempts made things worse, code is tangled with fixes patching fixes, it's no longer clear what the original problem was. → The session itself is polluted; more prompting compounds the damage. Response: Hard reset (Step 3B).
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.
- 1mo ago Changed e0ef2f175af3
- 1mo ago First seen · 81 lines · 59 tokens per session scan A b7ee760ec844
dredge is a skill published in the GitHub repository partikularwaters/Banka (3 stars, last pushed 1mo ago), licensed MIT. It adds 59 tokens to every session and 1,212 once invoked, about $0.0002 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 2026-08-31.
Other skills, from other repositories
debug
Investigate a failure, regression or incorrect result in an integration or application, even before it can be reproduced. Use for diagnosis and repair; follow incident authority for live mitigation.
rescue
Triage an outage, loss of stakeholder trust or a failing engagement direction. Use for urgent recovery; distinguish diagnosis from authorized production mitigation.
python-performance-optimization
Profile and optimize Python code using cProfile, memory profilers, and performance best practices.
prove-checks
Prove a passing check was capable of failing before recording it as evidence. Use when a test, CI job, build-and-diff, smoke test or rehearsal comes back green and that green is about to be treated as proof - especially when the check depends on a setup mutation (a sed/awk rewrite, an env var, a secret, a fixture…
logging-patterns
Logging conventions -- level usage, formatting style, structured output.
8051-mcu-debug
Use when debugging 8051-compatible microcontrollers, 51 MCU firmware, STC download issues, Keil C51 projects, interrupts, timers, UART, or startup failures.