llm cost optimization audit
Audit LLM spend to find recoverable waste in prompt caching, model routing, and token usage.
New: UPI payments are live. Buyers in India can now pay for skills with UPI in INR -> Browse skills
THE AGENSI STORE
146 skills found
Audit LLM spend to find recoverable waste in prompt caching, model routing, and token usage.
by Timoranjes
context-window-manager-pro
Audit cookie consent flows and manage DSAR workflows against active GDPR and ePrivacy standards.
by Timoranjes
A systematic debugging framework for AI agents to resolve edit loops, instruction decay, and context drift in production. Helps agents self-diagnose failure patterns and recover autonomously.
Audit and design behavioral-based moderation systems for communities to stop AI spam without relying on text detectors.
by heyhridyansh
A rigorous reviewer that audits implementation plans for architectural risks, scope creep, and safety before you code.
Audit email infrastructure against 2024/2025 Gmail, Yahoo, and Outlook bulk-sender compliance requirements.
by YK
Prevents AI agents from executing destructive Git or filesystem commands without explicit confirmation and impact analysis.
by Md Tamim
A systematic diagnostic framework for debugging AI pipelines, prompt drift, RAG failures, and agent orchestrations.
by heyhridyansh
Audits AI agents and automated workflows for infinite loops, cost spikes, and runaway execution risks.
by nowrich
Professional prompt auditing skill that stress-tests LLM instructions across 8 axes with scored verdicts and diffs.
by Vivek K
Designs a scored eval suite for an AI agent or SKILL.md skill, measures trigger accuracy, catches regressions against a baseline, and issues a PASS / PASS-WITH-WARNINGS / FAIL release gate.
by heyhridyansh
Diagnose RAG hallucinations, retrieval failures, and citation errors with a structured root-cause audit.
by Jiangdongbo
Turns vague requirements into three opinionated, costed architecture plans with an executive business summary.
by nowrich
Convert raw prompts into typed, variable-driven templates with built-in hallucination guardrails.
by UNIQUEBEE
Architects technical Multi-Agent System (MAS) blueprints and SOPs for scaling e-commerce operations.
by heyhridyansh
Evaluates dependency and framework upgrades for breaking changes, peer conflicts, and deployment risks.
by heyhridyansh
Architects deterministic, isolated, and privacy-safe test data systems to eliminate flakes and leaks.
by Vivek K
Turn messy CI/CD, Docker, and Kubernetes failure logs into an evidence-backed Root Cause Analysis and a safe, human-run recovery plan.
by UNIQUEBEE
Architects production-ready backend automation libraries and IaC for SaaS operations.
by Vivek K
Audits a repository for exposed secrets and vulnerable dependencies with threat-model context, false-positive filtering, secret lifecycle, attack-path analysis, remediation, and a CRITICAL-EXPOSURE → CLEAN maturity verdict — without ever printing full secret values.
by nowrich
Audit, compress, and prove instruction-preservation for verbose AI prompts to reduce API token costs.
FPGA reset architecture as an area decision: why one wrong reset turns 4 BRAMs into 73,728 flip-flops, and the hybrid reset pattern that is always safe.
by Vivek K
Designs multi-system enterprise automation the right way: what to automate vs keep manual, build vs buy vs orchestrate, idempotency and HITL gates, security, SLA/SLOs, safe rollout, and ROI — without running the workflow.