Find ruflo plugins suited to your current task, workflow, and installed MCP tools, which connect agents to external capabilities. Browse recommendations, search by need, and review what each candidate adds before choosing plugins.
Break down a Claude Code session's costs by assistant message to find expensive responses. Compare input, output, cache writes, cache reads, and models, with message cost percentiles providing context for investigating unusually costly activity.
Check Ruflo's health and summarize system diagnostics, MCP server state, and active agents. Routine checks leave the installation unchanged; an explicit repair request runs the repair command and reports the resulting changes and status.
Start a named browser session and preserve its initial screen, page structure, and action history. Keep a trace of site exploration and establish a reusable session for later extraction, form filling, or browser testing.
Fill web forms from a JSON list of field names and values, using saved templates or page labels to locate inputs. Optionally submit the form, save new field mappings, and redact values from recorded actions.
Extract text, tables, and attributes from rendered web pages into structured data. Check extracted content for personal information and malicious instructions, and save reusable extraction templates for pages on the same host.
Apply simple code changes across files without calling an LLM, using agent-booster's local transformation engine. The skill checks a confidence score before saving results and records outcomes for cost tracking and future routing decisions.
Compare recorded agent spending with a project budget and receive alerts at four usage thresholds. Review remaining funds and recommended actions, or use the check to block agent launches when the budget is exhausted.
Compare two saved cost-summary JSON snapshots to identify changes in total spending, sessions, and model costs. Set limits on spending increases or token growth to flag cost regressions during pull request checks.
Organize an AI team's work into small tasks with one owner and one deliverable each. Record dependencies, check for duplicate tasks, and update progress so the team's recorded task state stays clear.
Check untrusted text before passing it to a model or executing it. Using AIDefence, it provides safety checks, detailed threat classifications, confidence scores, and scanning for attempts to override instructions or extract data.
Compare actual AI session costs with estimates for using only Haiku, Sonnet, or Opus. The skill reports savings amounts and percentages, helping teams assess model selection costs while leaving differences in answer quality and retries outside the calculation.
Identify individual sessions with unusually high or low spending using the median and MAD, a measure of spread around it. Filter by time period, adjust detection thresholds, and optionally return a failure exit code for automated checks.
Summarize spending events for each connected peer over the past hour, day, and week, and flag peers above a configured suspension threshold. This consumer-side script depends on upstream event publishing that the source describes as not yet implemented.
Estimate future spending from recent usage costs, with projections for 7, 30, 90, and 365 days. When a budget is configured, see how many days remain before spending reaches key budget thresholds.
Review estimated spending, token usage, and cache hit ratios across Claude Code, Codex, and Grok using local logs. Filter by time, project, or provider, while keeping currencies separate and identifying models without pricing.
Run a prepared set of test cases to check agent-booster success rates and routing decisions. Optionally compare results with Gemini or Claude models, then save detailed measurements and summaries as JSON records for later verification.
Analyze Claude Code and Codex usage logs to identify spending drivers, including cache usage, model choices, reasoning, and long sessions. Receive evidence-backed recommendations and hypothetical savings estimates to help decide which changes to try.
Dry-run one cycle of a sales agent workflow without calling language models. Validate the template and agent types, check budget reservations, generate agent prompts, and record a structured status summary for downstream automation.
Generate a spending report from recorded AI usage, with token counts and USD costs grouped by model, agent, and processing tier. Check budget utilization, remaining funds, and alerts for a selected period.
Assess tasks before execution to identify transformations eligible to skip LLM calls. Report the eligible share, recommended models for remaining tasks, and estimated upper costs, then save the routing results for later cost reports.
Export recorded AI usage costs and budgets as Prometheus metrics or JSON. Feed external dashboards with totals, model tier breakdowns, session costs, and budget utilization, or send the same data to a webhook endpoint.
Track documented code fixes across releases with a signed manifest, a list of fixes and their identifying text. Verify that markers remain in source files, inspect file changes, and use recorded history to identify commits associated with regressions.
Run one cycle of autonomous task work in Claude Code: check progress, select and execute the next action, log the work, and schedule another cycle. Stop when all tasks are complete or the iteration limit is reached.
Is this your account? Link GitHub to bring it onto your creator page (Japanese)