ork-meter

Plugin ork-meter from marketplace edos in EDOS-Engineering/ork-meter-claude. Commands read from the repo's manifest — 5 days ago.

Skills (14)

• audit

  Audit what loads into every Claude Code session in this project - CLAUDE.md and rules files with line counts and estimated tokens, configured MCP servers (names only), and a ranked list of the largest always-loaded items. Use when the person asks to audit their setup, why sessions start large, or wh

• cache

  Report where this project's Claude Code tokens went - measured cache writes, cache reads and output from the local transcripts - and which prompt-cache misses re-sent whole conversations, by cause (idle past the cache lifetime, model or effort switch, compaction, tool or MCP changes). Use when the p

• capture

  Capture a selected noninteractive command before native error rendering can truncate its diagnostics. Preserve original streams and status for exact recovery without rerunning.

• claim

  Finish connecting with the approval code Ork Meter showed in a browser on another computer, after /ork-meter:connect offered the terminal approval.

• connect

  Connect once with browser sign-in and consent, then capture token counters automatically in normal Claude Code sessions. Works when the browser is on another computer too.

• disconnect

  Stop automatic capture, remove owned Claude settings and local credentials, and open the saved connection's remote revocation review.

• explore

  Ranked workspace search plus multi-file reads in one read-only call.

• lean

  Install, remove or check the opt-in lean main agent (fewer turns, targeted reads) for this project. Use when the person asks for the lean agent.

• map

  Generate or refresh a compact local map of this project (layout with sizes, entry points, build and test commands, manifests, conventions, top-level Python symbols) and store it as an on-demand project skill that Claude loads only when it needs orientation. Use when the person asks for a project map

• recover

  Recover omitted Ork Meter tool output without rerunning its command. Search a saved result or expand exact lines, with full original output always available.

• savings

  Show what the Ork Meter token-trim hooks did this session - interventions, bytes avoided, estimated tokens avoided, and the session's measured token usage from its transcript, each labelled. Use when the person asks about savings, trimmed output, tokens avoided, or what the trim hooks did.

• source

  Retrieve Python outlines, exact symbols, or bounded batches of source ranges and literal searches with content hashes. Other languages support exact ranges.

• status

  Check the saved connection, automatic receiver and hosted token-export receipts. No reference ID is needed.

• statusline

  Print the settings.json snippet that turns on the Ork Meter status line - context size, estimated cost of the current turn and of the session at API list prices, and whether the prompt cache has likely expired. Use when the person asks for a cost or context status line.

Hooks (20)

• SessionStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/session-start.sh" "${CLAUDE_PLUGIN_DATA}"

• SessionStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/python-check.sh"

• SessionStart

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" trim-session.py

• PreToolUse

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" read-guard.py

• PreToolUse

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" write-guard.py

• PreToolUse

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" polling-completion.py

• PostToolUse

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" trim-tool-output.py

• PostToolUse

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" dedupe-reads.py

• PostToolUse

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" compact-records.py

• PostToolUse

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" write-guard.py

• PostToolUse

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" polling-completion.py

• PostToolUseFailure

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" loop-guard.py

• PostToolUseFailure

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" trim-tool-output.py

• PostToolUseFailure

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" polling-completion.py

• SubagentStop

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" agent-stop.py

• UserPromptSubmit

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" context-nudge.py

• UserPromptSubmit

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" stale-context.py

• UserPromptSubmit

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" polling-completion.py

• SessionEnd

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" session-summary.py

• PostToolBatch

  bash "${CLAUDE_PLUGIN_ROOT}/hooks/py.sh" polling-completion.py

Similar plugins