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/backchainai/backchain-plugins/auditnpx skills add backchainai/backchain-plugins --skill auditgit clone --depth 1 https://github.com/backchainai/backchain-pluginsWhat 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 | $0.00149 | $0.02432 |
| Opus 5 | $0.00075 | $0.01216 |
| Sonnet 5 | $0.00030 | $0.00486 |
| Haiku 4.5 | $0.00015 | $0.00243 |
Grade A, and why
audit 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 2d 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 — 148 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AI-Slop Content Audit
Audit the content described in $ARGUMENTS and return a senior-reviewer report. You are running in a forked context with no access to upstream conversation; rely only on the arguments and the bundled research file.
Inputs
The operator's request will name (or contain inline):
- The content to audit (verbatim text, a file path, or a URL).
- The audience and medium (for example, "AI engineers reading a staff-engineer candidate's portfolio," "SMB clients reading a consulting proposal").
- Optional: the author's claimed voice or role.
If the content is a file path, read it. If it is a URL, fetch it. If any input is missing, ask once for it, then proceed.
Detection framework
Apply these six categories in order. Note specific spans for each tell.
1. Lexical tells
Focal words whose frequency jumped in post-ChatGPT text (Juzek & Ward, COLING 2025; arxiv:2412.11385): delve, delves, delving, showcasing, showcases, boasts, underscores, underscoring, underscore, comprehending, intricacies, intricate, surpassing, garnered, emphasizing, realm, groundbreaking, advancements, aligns.
GPT-4o lexical overuse relative to human text (Reinhart et al., PNAS 2025; arxiv:2410.16107): camaraderie, tapestry, intricate, underscore, unspoken, amidst, palpable, solace, fleeting, unravel. (intricate and underscore appear in both lists.)
Common LLM tells not tied to a single study (cite as general slop markers, not to a paper): robust, leverage, seamless, navigate, enhance, facilitate, landscape, elevate, empower, unlock, unparalleled, foster.
Phrases that recur in LLM output: "it is important to note," "in today's fast-paced world," "navigate the landscape," "dive into," "at its core," "in the realm of," "unlock the potential," "here's what you need to know."
Flag any cluster of three or more within a short passage, or use of a signature token where a simpler word would serve.
2. Structural and syntactic patterns
- Tricolon overuse: "X, Y, and Z" constructions. Eight or more in a piece shifts probability toward AI authorship. Reinhart et al. report phrasal coordination at 1.9x human rate in GPT-4o (
arxiv:2410.16107); see${CLAUDE_SKILL_DIR}/references/research.md. - Uniform sentence length: LLM output concentrates in the 10 to 30 token band; human writing scatters with short fragments and long runs. Muñoz-Ortiz et al. report this sentence-length clustering (AI Review 2024); see references file.
- Ascending parallelism: tricolons where each item grows longer by design.
- Pronoun suppression: low "I," "we," "you" density when the topic is personal.
- Passive voice and nominalization: "Findings suggest" instead of "I found."
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.
- 2d ago First seen · 148 lines · 0 tokens per session scan A 93d2d908bcff
audit is a skill published in the GitHub repository backchainai/backchain-plugins (4 stars, last pushed 27d ago), licensed Apache-2.0. It adds 149 tokens to every session and 2,432 once invoked, about $0.0007 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 2026-08-31.
Other skills, from other repositories
kayba-stage-3-metrics
Define metrics from Kayba insights, implement them as Python measurement code, run against traces, and iterate until the metrics are clean and meaningful. Trigger when the user says "run stage 3", "define metrics", "build metrics", "compute baselines", or when invoked by the kayba-pipeline orchestrator. Requires…
kayba-stage-5-action-plan
Triage each insight into discard/code-fix/prompt-fix and produce a prioritized action plan with specific recommendations. Trigger when the user says "run stage 5", "make action plan", "triage skills", or when invoked by the kayba-pipeline orchestrator. Requires eval outputs from stages 1-4.
kayba-stage-6-hitl
Human-In-The-Loop gate that presents the action plan with full context, collects an informed approval/modification/rejection decision, and records the outcome. Trigger when the user says "run stage 6", "HITL review", "approve action plan", or when invoked by the kayba-pipeline orchestrator. Requires eval/actionplan.md…
kayba-pipeline
End-to-end agent evaluation and improvement pipeline. Takes a traces folder and optional HITL flag, then orchestrates sub-agents through 7 stages — each stage is its own skill invoked by a dedicated sub-agent. Trigger when the user says "run the pipeline", "kayba pipeline", "evaluate and fix", "full eval", "analyze…
kayba-stage-2-domain-context
Gather domain context about the repository and agent — system prompt, tool definitions, domain docs, and behavior patterns from traces. Trigger when the user says "run stage 2", "gather context", "domain context", or when invoked by the kayba-pipeline orchestrator.
kayba-stage-7-fixer
Implement the approved fixes from the action plan and log all changes. Trigger when the user says "run stage 7", "implement fixes", "apply action plan", or when invoked by the kayba-pipeline orchestrator. Requires eval/actionplan.md to exist.