@kurrent/kapacitor 0.5.9 → 0.5.10

package/kapacitor/.claude-plugin/plugin.json

@@ -1,5 +1,5 @@
 {
   "name": "kapacitor",
-  "version": "0.5.9",
+  "version": "0.5.10",
   "description": "Records and visualizes Claude Code sessions via kapacitor CLI hooks"
 }

package/kapacitor/.codex-plugin/plugin.json

@@ -2,6 +2,5 @@
   "name": "kapacitor",
   "version": "1.6.0",
   "description": "Records and visualizes Codex CLI sessions via kapacitor CLI hooks",
-  "skills": "./codex-skills/",
   "mcpServers": "./.codex-mcp.json"
 }

package/kapacitor/README.md

@@ -46,15 +46,15 @@ PR review context tools, available automatically when the agent is on a branch w
 
 Each hook pipes its JSON payload through the `kapacitor` CLI, which enriches it with git/PR info and forwards it to the server. A background watcher process streams transcript lines in real time.
 
-**Skills** — Slash commands for reviewing recorded sessions. Available on both harnesses (Claude reads `skills/`, Codex reads `codex-skills/`):
+**Skills** — Slash commands for reviewing recorded sessions. Claude reads `skills/` (unprefixed folder names); Codex and Cursor read `~/.agents/skills/` (`kapacitor-*` prefixed folder names):
 
-- `kapacitor-recap` — Retrieve a structured summary of a session (user prompts, assistant responses, plans, file changes)
-- `kapacitor-errors` — Extract tool call errors from a session for post-session review and pattern detection
-- `kapacitor-validate-plan` — Verify that all planned items were completed
-- `kapacitor-disable` — Stop recording and delete all server data for the current session
-- `kapacitor-hide` — Hide the current session (owner-only visibility)
+- `recap` / `kapacitor-recap` — Retrieve a structured summary of a session (user prompts, assistant responses, plans, file changes)
+- `errors` / `kapacitor-errors` — Extract tool call errors from a session for post-session review and pattern detection
+- `validate-plan` / `kapacitor-validate-plan` — Verify that all planned items were completed
+- `disable` / `kapacitor-disable` — Stop recording and delete all server data for the current session
+- `hide` / `kapacitor-hide` — Hide the current session (owner-only visibility)
 
-In Claude they're invoked as `/kapacitor:session-recap`, `/kapacitor:session-errors`, etc.
+In Claude they're invoked as `/kapacitor:recap`, `/kapacitor:errors`, etc.
 
 ## Prerequisites
 
@@ -67,8 +67,9 @@ In Claude they're invoked as `/kapacitor:session-recap`, `/kapacitor:session-err
 
 ```bash
 kapacitor plugin install            # Claude Code, user-wide
-kapacitor plugin install --codex    # Codex CLI, user-wide
-kapacitor plugin install --project  # current project only
+kapacitor plugin install --codex    # Codex CLI, user-wide (hooks + agent skills)
+kapacitor plugin install --skills   # Agent skills only (~/.agents/skills/), no Codex hooks
+kapacitor plugin install --project  # current project only (hooks scope; skills always user-wide)
 ```
 
 ### Option B: Interactive plugin manager
@@ -76,6 +77,8 @@ kapacitor plugin install --project  # current project only
 - Claude Code: run `/plugin` inside a session and browse the **Installed** tab.
 - Codex CLI: `codex plugin marketplace add kurrent-io/kapacitor-cli` then enable from the marketplace.
 
+> **Note:** Codex's native plugin loader installs the MCP servers only. To get hooks and agent skills, additionally run `kapacitor plugin install --codex`.
+
 ### Option C: Settings file (manual)
 
 Add to `.claude/settings.local.json` or `~/.claude/settings.json`:
@@ -117,32 +120,21 @@ kapacitor/
     plugin.json          — Claude manifest (name, version, description)
     marketplace.json     — Marketplace manifest for plugin discovery
   .codex-plugin/
-    plugin.json          — Codex manifest (refs ./.codex-mcp.json and ./codex-skills/)
+    plugin.json          — Codex manifest (refs ./.codex-mcp.json)
   .mcp.json              — Claude MCP servers (camelCase mcpServers shape)
   .codex-mcp.json        — Codex MCP servers (snake_case mcp_servers shape)
   hooks/
     hooks.json           — Hook definitions for all Claude lifecycle events
   skills/
-    session-recap/
-      SKILL.md           — /kapacitor:session-recap skill (Claude)
-    session-errors/
+    recap/
+      SKILL.md           — /kapacitor:recap skill (Claude)
+    errors/
       SKILL.md
     validate-plan/
       SKILL.md
-    session-disable/
-      SKILL.md
-    session-hide/
-      SKILL.md
-  codex-skills/
-    kapacitor-recap/
-      SKILL.md           — same skill, Codex variant
-    kapacitor-errors/
-      SKILL.md
-    kapacitor-validate-plan/
-      SKILL.md
-    kapacitor-disable/
+    disable/
       SKILL.md
-    kapacitor-hide/
+    hide/
       SKILL.md
 ```
 

package/kapacitor/skills/{session-disable → disable}/SKILL.md

@@ -1,5 +1,5 @@
 ---
-name: session-disable
+name: disable
 description: >-
   This skill should be used when the user asks to "disable recording",
   "stop recording", "delete this session", "don't record this",
@@ -28,7 +28,7 @@ This will:
 
 ## Requirements
 
-- Must be run inside an active Claude Code session (`KAPACITOR_SESSION_ID` must be set)
+- `kapacitor disable` resolves the current session id from the environment when the host agent CLI exposes one. If no session id is available, pass it explicitly: `kapacitor disable <sessionId>`.
 - The kapacitor CLI must be on PATH
 
 ## Notes

package/kapacitor/skills/{session-errors → errors}/SKILL.md

@@ -1,5 +1,5 @@
 ---
-name: session-errors
+name: errors
 description: >-
   This skill should be used when the user asks to "show errors", "extract errors",
   "what went wrong", "find tool errors", "review errors from session",
@@ -11,11 +11,11 @@ description: >-
 
 # Session Errors
 
-Extract tool call errors from a Claude Code session recorded by Kurrent Capacitor. The output lists each failed tool call — bash commands, file reads/writes, agent delegations, etc. — along with the error message and the tool that caused it.
+Extract tool call errors from an agent session recorded by Kurrent Capacitor. The output lists each failed tool call — bash commands, file reads/writes, agent delegations, etc. — along with the error message and the tool that caused it.
 
 ## Usage
 
-Run `kapacitor errors` via the Bash tool. The session ID is automatically set by the `KAPACITOR_SESSION_ID` environment variable (persisted at session start).
+Run `kapacitor errors` via the Bash tool. `kapacitor errors` resolves the current session id from the environment when the host agent CLI exposes one. If no session id is available, pass it explicitly: `kapacitor errors <sessionId>`.
 
 ```bash
 # Errors from the current session
@@ -47,7 +47,7 @@ When using `--chain`, errors from all sessions in the continuation chain are inc
 
 ## Practical Applications
 
-- **End-of-session review** — run after finishing a session to identify recurring mistakes and update CLAUDE.md with avoidance rules
+- **End-of-session review** — run after finishing a session to identify recurring mistakes and add avoidance rules to the project's agent instructions file (e.g. `AGENTS.md`).
 - **Debugging** — quickly find what went wrong in a session without scrolling through the full timeline
 - **Pattern detection** — use `--chain` across a multi-session task to spot repeated error patterns (e.g., wrong file paths, incorrect API usage)
 
@@ -58,7 +58,7 @@ The `KAPACITOR_URL` environment variable overrides the default server URL (`http
 ## Tips
 
 - After extracting errors, look for patterns: the same tool failing repeatedly, or the same type of mistake across sessions.
-- Propose concrete avoidance rules based on the errors found — these can be added to the project's CLAUDE.md.
+- Propose concrete avoidance rules based on the errors found — these can be added to the project's agent instructions file.
 - The `kapacitor` CLI must be available on PATH (typically installed at `~/.local/bin/kapacitor`).
 
 ## Error Handling

package/kapacitor/skills/{session-hide → hide}/SKILL.md

@@ -1,5 +1,5 @@
 ---
-name: session-hide
+name: hide
 description: >-
   This skill should be used when the user asks to "hide this session",
   "make this private", "hide session", "owner only", "make private",
@@ -24,7 +24,7 @@ This sets the session visibility to "none" (owner-only). Other users will no lon
 
 ## Requirements
 
-- Must be run inside an active Claude Code session (`KAPACITOR_SESSION_ID` must be set)
+- `kapacitor hide` resolves the current session id from the environment when the host agent CLI exposes one. If no session id is available, pass it explicitly: `kapacitor hide <sessionId>`.
 - The kapacitor CLI must be on PATH
 
 ## Notes

package/kapacitor/skills/{session-recap → recap}/SKILL.md

@@ -1,5 +1,5 @@
 ---
-name: session-recap
+name: recap
 description: >-
   This skill should be used when the user asks to "read a previous session",
   "get session history", "recap session", "what happened in session X",
@@ -44,7 +44,7 @@ kapacitor recap --full <sessionId>
 kapacitor recap --chain <sessionId>
 ```
 
-The session ID is automatically set by the `KAPACITOR_SESSION_ID` environment variable (persisted at session start). You can pass an explicit ID to review a different session.
+`kapacitor recap` resolves the current session id from the environment when the host agent CLI exposes one. If no session id is available, pass it explicitly: `kapacitor recap <sessionId>`.
 
 ## Repository Recap (`--repo`)
 
@@ -72,7 +72,7 @@ If no summary is available (e.g., active session), a hint is shown to use `--ful
 The complete transcript with these section types:
 
 - **`## User Prompt`** — what the user asked
-- **`## Assistant`** — Claude's text responses
+- **`## Assistant`** — Agent text responses
 - **`## Plan`** — plans that were created
 - **`## Write <path>`** — files that were created (with syntax-highlighted content)
 - **`## Edit <path>`** — files that were edited (with diff content)

package/kapacitor/skills/validate-plan/SKILL.md

@@ -9,11 +9,11 @@ description: >-
 
 # Validate Plan
 
-Verify that all items in the current session's plan have been completed. Plans come from either a continuation (`SessionStarted.planContent`) or an in-session `ExitPlanMode` write to `~/.claude/plans/`.
+Verify that all items in the current session's plan have been completed. Plans come from either a session continuation (`SessionStarted.planContent`) or an in-session plan recorded by the kapacitor hooks.
 
 ## Usage
 
-Run `kapacitor validate-plan` via the Bash tool. The session ID is automatically set by the `KAPACITOR_SESSION_ID` environment variable (persisted at session start).
+Run `kapacitor validate-plan` via the Bash tool. `kapacitor validate-plan` resolves the current session id from the environment when the host agent CLI exposes one. If no session id is available, pass it explicitly: `kapacitor validate-plan <sessionId>`.
 
 ```bash
 # Validate the current session's plan
@@ -44,7 +44,7 @@ The command outputs three sections:
 
 If the output says "No plan found for this session", inform the user that no plan was detected for this session. A plan is only present when:
 - The session continued from a previous session that had a plan (`planContent`)
-- The session used `ExitPlanMode` to create a plan file (written to `~/.claude/plans/`)
+- The session recorded an in-session plan via the kapacitor hooks.
 
 ## Environment
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@kurrent/kapacitor",
-  "version": "0.5.9",
+  "version": "0.5.10",
   "description": "CLI companion for Kurrent Capacitor — records and visualizes Claude Code sessions",
   "license": "UNLICENSED",
   "repository": {
@@ -11,12 +11,12 @@
     "kapacitor": "bin/kapacitor.js"
   },
   "optionalDependencies": {
-    "@kurrent/kapacitor-darwin-arm64": "0.5.9",
-    "@kurrent/kapacitor-linux-x64": "0.5.9",
-    "@kurrent/kapacitor-linux-arm64": "0.5.9",
-    "@kurrent/kapacitor-linux-musl-x64": "0.5.9",
-    "@kurrent/kapacitor-linux-musl-arm64": "0.5.9",
-    "@kurrent/kapacitor-win-x64": "0.5.9"
+    "@kurrent/kapacitor-darwin-arm64": "0.5.10",
+    "@kurrent/kapacitor-linux-x64": "0.5.10",
+    "@kurrent/kapacitor-linux-arm64": "0.5.10",
+    "@kurrent/kapacitor-linux-musl-x64": "0.5.10",
+    "@kurrent/kapacitor-linux-musl-arm64": "0.5.10",
+    "@kurrent/kapacitor-win-x64": "0.5.10"
   },
   "files": [
     "bin/",

package/kapacitor/codex-skills/kapacitor-disable/SKILL.md

@@ -1,45 +0,0 @@
----
-name: kapacitor-disable
-description: >-
-  This skill should be used when the user asks to "disable recording",
-  "stop recording", "delete this session", "don't record this",
-  "remove my session data", "erase this session", "turn off tracking",
-  "stop tracking", "disable kapacitor", "remove session",
-  or wants to stop the current session from being recorded.
-  Stops the watcher, prevents future hooks, and deletes server data.
----
-
-# Kapacitor Disable
-
-Stop recording the current session and delete all data from the server.
-
-## Usage
-
-```bash
-# Active session — auto-resolved from CODEX_THREAD_ID (Codex 0.81+)
-kapacitor disable
-
-# Specific session by ID (from `kapacitor recap --repo`)
-kapacitor disable <sessionId>
-```
-
-This will:
-1. Stop the transcript watcher process
-2. Prevent all future hook events from being sent for this session
-3. Delete all session data from the server (event streams and read models)
-
-## Requirements
-
-Run inside an active Codex CLI session (0.81+) — the active session is auto-resolved from `CODEX_THREAD_ID`. To act on a different session, pass its ID explicitly. Use `kapacitor recap --repo` to list recent session IDs in this repository.
-
-The `kapacitor` CLI must be available on PATH (typically installed at `~/.local/bin/kapacitor`).
-
-## Notes
-
-- This action is irreversible — all session data will be permanently deleted from the server.
-- The local transcript file (Codex rollout `.jsonl`) is not affected — only server-side data is removed.
-- Subsequent hooks (session-end, etc.) will be silently skipped.
-
-## Environment
-
-`KAPACITOR_URL` overrides the default server URL (`http://localhost:5108`).

package/kapacitor/codex-skills/kapacitor-errors/SKILL.md

@@ -1,70 +0,0 @@
----
-name: kapacitor-errors
-description: >-
-  Use when the user asks to "show errors", "extract errors", "what went wrong",
-  "find tool errors", "review errors from session", "check session errors",
-  "list failures", "what failed in session X", "error report", "show mistakes",
-  or wants to review tool call errors from a recorded session. Extracts tool
-  errors via the `kapacitor errors` CLI.
----
-
-# Kapacitor Errors
-
-Extract tool call errors from a session recorded by Kurrent Capacitor. The output lists each failed tool call — shell commands, file reads/writes, agent delegations, etc. — along with the error message and the tool that caused it.
-
-Codex CLI 0.81+ exports `CODEX_THREAD_ID`; `kapacitor errors` uses it the same way it uses `KAPACITOR_SESSION_ID` for Claude. No args needed for the current session.
-
-## Usage
-
-```bash
-# Current session — auto-resolved from CODEX_THREAD_ID
-kapacitor errors
-
-# Errors from the full continuation chain of the current session
-kapacitor errors --chain
-
-# Errors from a specific session
-kapacitor errors <sessionId>
-
-# Errors from the full continuation chain starting at a session
-kapacitor errors --chain <sessionId>
-```
-
-## Output Format
-
-Each error is printed as a block with:
-
-- **Session ID** and optional **agent ID** (if the error occurred in a subagent).
-- **Event number** and **timestamp**.
-- **Tool name** — the tool that failed (e.g., Bash, Read, Edit, Write, Grep, Glob).
-- **Error message** — the error output or failure reason.
-
-When using `--chain`, errors from all sessions in the continuation chain are included, ordered chronologically.
-
-## When to Use Each Flag
-
-- **No flag** (`kapacitor errors`) — reviewing errors from the current session.
-- **`<sessionId>`** — reviewing errors from a specific session (e.g. from `kapacitor recap --repo`).
-- **`--chain`** — reviewing errors across a full task that spanned multiple sessions.
-
-## Practical Applications
-
-- **Post-mortem review** — after finishing a session, identify recurring mistakes and update project rules (CLAUDE.md / AGENTS.md) with avoidance guidance.
-- **Debugging** — quickly find what went wrong in a session without scrolling through the full timeline.
-- **Pattern detection** — use `--chain` across a multi-session task to spot repeated error patterns (e.g., wrong file paths, incorrect API usage).
-
-## Environment
-
-`KAPACITOR_URL` overrides the default server URL (`http://localhost:5108`).
-
-## Tips
-
-- After extracting errors, look for patterns: the same tool failing repeatedly, or the same type of mistake across sessions.
-- Propose concrete avoidance rules based on the errors found — these can be added to AGENTS.md / CLAUDE.md.
-- The `kapacitor` CLI must be available on PATH (typically installed at `~/.local/bin/kapacitor`).
-
-## Error Handling
-
-- If the session is not found, the command prints an error and exits with code 1.
-- If no errors are found, the command prints "No errors found." — this is a good outcome.
-- If the Kurrent Capacitor server is unreachable, the command prints an HTTP error. Ensure the server is running.

package/kapacitor/codex-skills/kapacitor-hide/SKILL.md

@@ -1,41 +0,0 @@
----
-name: kapacitor-hide
-description: >-
-  This skill should be used when the user asks to "hide this session",
-  "make this private", "hide session", "owner only", "make private",
-  "hide from others", "set private", "don't show this session",
-  or wants to change the current session visibility to owner-only.
-  Sets session visibility so only the owner can see it.
----
-
-# Kapacitor Hide
-
-Hide the current session so only you (the owner) can see it.
-
-## Usage
-
-```bash
-# Active session — auto-resolved from CODEX_THREAD_ID (Codex 0.81+)
-kapacitor hide
-
-# Specific session by ID (from `kapacitor recap --repo`)
-kapacitor hide <sessionId>
-```
-
-This sets the session visibility to "none" (owner-only). Other users will no longer see this session in the Capacitor dashboard.
-
-## Requirements
-
-Run inside an active Codex CLI session (0.81+) — the active session is auto-resolved from `CODEX_THREAD_ID`. To act on a different session, pass its ID explicitly. Use `kapacitor recap --repo` to list recent session IDs in this repository.
-
-The `kapacitor` CLI must be available on PATH (typically installed at `~/.local/bin/kapacitor`).
-
-## Notes
-
-- This is reversible — visibility can be changed back via the Capacitor dashboard.
-- The session data remains on the server, just hidden from other users.
-- Unlike `kapacitor disable`, recording continues normally.
-
-## Environment
-
-`KAPACITOR_URL` overrides the default server URL (`http://localhost:5108`).

package/kapacitor/codex-skills/kapacitor-recap/SKILL.md

@@ -1,95 +0,0 @@
----
-name: kapacitor-recap
-description: >-
-  Use when the user asks to "read a previous session", "get session history",
-  "recap session", "what happened in session X", "load context from a previous
-  session", "continue from session", "what did we do last time", "catch me up",
-  "summarize session", "what have we been working on", "recent changes",
-  "recent sessions", or references prior work in this repo. Retrieves session
-  summaries recorded by Kurrent Capacitor via the `kapacitor recap` CLI.
----
-
-# Kapacitor Recap
-
-Retrieve session history recorded by Kurrent Capacitor. Codex CLI 0.81+ exports `CODEX_THREAD_ID`; `kapacitor recap` uses it the same way it uses `KAPACITOR_SESSION_ID` for Claude. No args needed for the current session.
-
-## Usage
-
-Run the `kapacitor recap` CLI via the shell. Do NOT call the HTTP API directly — the CLI handles formatting, error handling, and server URL resolution.
-
-```bash
-# Current session — auto-resolved from CODEX_THREAD_ID
-kapacitor recap
-
-# Recent session summaries for the current repository (discovery)
-kapacitor recap --repo
-
-# A specific session by ID (from --repo output)
-kapacitor recap <sessionId>
-
-# Full transcript for a specific session
-kapacitor recap --full <sessionId>
-
-# Full continuation chain (all linked sessions, oldest first)
-kapacitor recap --chain <sessionId>
-
-# Both: full transcript across all chained sessions
-kapacitor recap --chain --full <sessionId>
-```
-
-## Default Output (Summary)
-
-Shows the plan (if any) and an AI-generated summary with:
-- **Context** — why the work was done.
-- **Key decisions** — trade-offs and design choices that matter for future work.
-- **Unfinished/Risks** — anything deferred or left incomplete.
-
-If no summary is available (e.g., active session), a hint is shown to use `--full`.
-
-## Repository Recap (`--repo`)
-
-Returns AI-generated summaries from the most recent ended sessions in the current git repository. Each entry includes the session title, date, summary, and the session ID needed to drill in further.
-
-**Use this for discovery — not as the default entry point:**
-- The user says "what have we been working on recently?"
-- The user references prior work ("recently we implemented X")
-- You need context about other sessions in this repo, not the current one
-
-**Progressive disclosure:** Start with `kapacitor recap` (current session) for in-session work. Switch to `--repo` when you need cross-session context. Drill into a specific summary with `kapacitor recap --full <sessionId>`.
-
-## Full Output (`--full`)
-
-The complete transcript with these section types:
-
-- **`## User Prompt`** — what the user asked.
-- **`## Assistant`** — text responses.
-- **`## Plan`** — plans that were created (Claude Code sessions only).
-- **`## Write <path>`** — files that were created (with syntax-highlighted content).
-- **`## Edit <path>`** — files that were edited (with diff content).
-
-When using `--chain`, sessions are separated by `# Session <id>` headers, and agent activity appears under `### Agent (<type>)` sub-headers.
-
-## When to Use Each Flag
-
-- **No flag** (`kapacitor recap`) — quick context on the current session.
-- **`--repo`** — recent session summaries across the repo.
-- **`<sessionId>`** — quick context on a specific session (from `--repo` output).
-- **`--full <sessionId>`** — when you need exact prompts, responses, or file contents.
-- **`--chain <sessionId>`** — understanding the full history of a task that spanned multiple sessions.
-- **`--chain --full <sessionId>`** — complete transcript across all continuations.
-
-## Environment
-
-`KAPACITOR_URL` overrides the default server URL (`http://localhost:5108`).
-
-## Tips
-
-- Start with `kapacitor recap` (no args) for the active session, then `--repo` for cross-session discovery.
-- Summarize key decisions and changes for the user rather than echoing the full recap output verbatim.
-- The `kapacitor` CLI must be available on PATH (typically installed at `~/.local/bin/kapacitor`).
-
-## Error Handling
-
-- If the session is not found, the command prints "Session not found" and exits with code 1.
-- If not in a git repository (for `--repo`), the command prints an error and exits with code 1.
-- If the Kurrent Capacitor server is unreachable, the command prints an HTTP error. Ensure the server is running.

package/kapacitor/codex-skills/kapacitor-validate-plan/SKILL.md

@@ -1,53 +0,0 @@
----
-name: kapacitor-validate-plan
-description: >-
-  This skill should be used when the user asks to "validate plan",
-  "verify plan", "check plan completion", "did I finish everything",
-  "is the plan done", "what's left to do", "validate my work",
-  or wants to verify that all planned items were completed.
----
-
-# Kapacitor Validate Plan
-
-Verify that all items in the current session's plan have been completed.
-
-## Usage
-
-```bash
-# Active session — auto-resolved from CODEX_THREAD_ID (Codex 0.81+)
-kapacitor validate-plan
-
-# Specific session by ID (from `kapacitor recap --repo`)
-kapacitor validate-plan <sessionId>
-```
-
-## What It Returns
-
-The command outputs three sections:
-
-- **`## Plan`** — the full plan text.
-- **`## What's Done`** — two sub-sections:
-  - **Summary** — AI-generated summary of what was accomplished (from `WhatsDoneGenerated` events, if available).
-  - **Details** — list of files created (`Write`) and modified (`Edit`) during the session.
-- **`## Instructions`** — asks you to compare the plan against the summary and file list.
-
-## What To Do With The Output
-
-1. Read the plan carefully and identify each distinct planned item.
-2. Compare each item against the summary and file list under "What's Done".
-3. If all items are complete, confirm to the user that the plan is fully implemented.
-4. If there are gaps, list the missing items and complete them now.
-
-## When No Plan Is Found
-
-If the output says "No plan found for this session", inform the user that no plan was detected for this session. Plans are recorded when a session emits an `ExitPlanMode`-style event; not every session has one.
-
-## Requirements
-
-Run inside an active Codex CLI session (0.81+) — the active session is auto-resolved from `CODEX_THREAD_ID`. To act on a different session, pass its ID explicitly.
-
-The `kapacitor` CLI must be available on PATH (typically installed at `~/.local/bin/kapacitor`).
-
-## Environment
-
-`KAPACITOR_URL` overrides the default server URL (`http://localhost:5108`).