notifyr: ntfy send/read client for agents

send (title/priority/tags), read (poll mode, greppable one-liners),
topics (known homelab buses from config). Config with homelab defaults
on first run. Unit tests against httptest; smoked against the real
ntfy on topic agent-tools-test. Spec: doc/tool-parity.md 3.3.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_011KikHkfCiC3yELbsMN8fT9

README.md

@@ -17,6 +17,15 @@ agent understands the result without opening an image viewer.
 Each tool has its own folder, its own README with the full format/CLI
 reference, its own tests and its own dev branch (`dev/<tool>`).
 
+## Planned: agent-capability & homelab-admin tools
+
+[`doc/tool-parity.md`](doc/tool-parity.md) compares Gemini CLI's
+built-in tools with Claude Code's, and specs the CLI tools that close
+the gaps (`notifyr`, `giteactl`, `waitfor`, `cronr`, `fleet`,
+`envaudit`, `reghelper`, `pagepub`, `nbcell`, `wtreectl`, `fanout`) so
+any agent gets the same capabilities via `run_shell_command`. Build
+order and rationale live there and in [`doc/plan.md`](doc/plan.md).
+
 ## Building
 
 ```bash

doc/plan.md

@@ -37,6 +37,26 @@ master/main. Only tools whose folders changed get rebuilt.
 - New tools: bitmap-font maker, sound-effect generator (sfxr-style text
   presets), tiled-map (.tmx) writer.
 
+### Agent-capability tools (see [tool-parity.md](tool-parity.md))
+
+Close the gap between Gemini CLI and Claude Code, plus homelab admin
+tools grounded in infra-Doc history. Suggested order:
+
+1. `notifyr` — send/read on the existing Pi5 ntfy bus (≈ PushNotification).
+2. `giteactl` — Gitea Actions runs, zst CI logs, `wait`/`wait-quiet`
+   build serialization, releases.
+3. `waitfor` — block until a condition holds (≈ Monitor);
+   `cronr` — systemd-user-timer scheduling (≈ CronCreate/List/Delete).
+4. `fleet` — one-shot homelab health snapshot via the read-only
+   claude-docker wrapper.
+5. `envaudit` — compose ↔ `.env` key/inline-secret audit;
+   `reghelper` — registry catalog + prune *plans*;
+   `pagepub` — publish HTML/MD report to a URL (≈ Artifact);
+   `nbcell` — Jupyter cell editing (≈ NotebookEdit);
+   `wtreectl` — disposable git worktrees.
+6. `fanout` — parallel headless-agent orchestration (≈ Workflow); only
+   on concrete need.
+
 ## Open questions
 
 - Versioned releases (`vX.Y.Z` tags per tool) on top of the rolling

doc/tool-parity.md

@@ -0,0 +1,297 @@
+# Tool parity: Google's agent vs Claude Code
+
+Goal: let Google's agent CLI do everything Claude Code can, by building
+the missing capabilities as small CLI tools in this repo. The agent
+calls them through its shell tool, so every tool follows the house
+rules: **text in, verifiable artifacts out**, single static Go binary,
+self-explanatory output.
+
+**Which Google agent?** Legacy `gemini-cli` is auth-dead for personal
+accounts (verified again 2026-08-05, see infra-Doc
+`hosts/brasse-linux01.md`). The real target is **`agy` (Antigravity
+CLI)** — live-tested in section 2, and its toolset differs from the
+old Gemini CLI docs. Section 1's table is kept for reference since
+Gemini CLI still exists in API-key mode.
+
+A design rule that fell out of the live test: **a dedicated binary
+beats ad-hoc shell because of approval prefixes.** agy (like Claude
+Code) allowlists commands by prefix — `notifyr …` can be approved once
+and forever, while every hand-rolled `for i in $(seq …); do curl …`
+loop is a unique string that needs fresh human approval. Small stable
+CLIs are therefore not just convenience: they are what makes
+unattended agent operation possible at all.
+
+Sources: Gemini CLI tools reference (<https://geminicli.com/docs/reference/tools/>),
+Claude Code's toolset as of 2026-08, live probing of `agy` 1.1.9.
+
+## 1. Already at parity — nothing to build
+
+| Capability | Claude Code | Gemini CLI |
+|---|---|---|
+| Read/write/edit files | `Read` / `Write` / `Edit` | `read_file` / `write_file` / `replace` |
+| Find files / search text / list dirs | `Glob` / `Grep` | `glob` / `grep_search` / `list_directory`, plus `read_many_files` |
+| Shell, incl. background processes | `Bash` (+ background tasks) | `run_shell_command` (+ background processes) |
+| Web | `WebFetch` / `WebSearch` | `web_fetch` / `google_web_search` |
+| Ask the user a structured question | `AskUserQuestion` | `ask_user` |
+| Plan mode | `EnterPlanMode` / `ExitPlanMode` | `enter_plan_mode` / `exit_plan_mode` |
+| Skills / slash commands | `Skill` (`.claude/skills`) | `activate_skill` (`.gemini/skills`) |
+| Persistent memory | file-based memory dir | `save_memory` (simpler, but exists) |
+| Todo/task tracking | `TaskCreate`/`TaskUpdate`/… | `write_todos`, `tracker_*` (experimental) |
+| MCP servers + resources | MCP tools, `ListMcpResources`/`ReadMcpResource` | MCP tools, `list_mcp_resources`/`read_mcp_resource` |
+| Subagents | `Agent` (background, custom types) | subagents (experimental) — weaker, see `fanout` below |
+
+Not worth replicating (harness-internal to Claude Code, no value as a
+CLI): `ToolSearch`, `EndConversation`, `ReportFindings`,
+`ShareOnboardingGuide`, `DesignSync`, remote cloud execution.
+
+## 2. Live test 2026-08-05: `agy` (Antigravity CLI 1.1.9)
+
+Tested interactively in a tmux session (Google AI Pro account, model
+Gemini 3.6 Flash). Its 19 built-in tools, self-enumerated:
+
+`ask_permission`, `ask_question`, `define_subagent`, `generate_image`,
+`grep_search`, `invoke_subagent`, `list_dir`, `list_permissions`,
+`manage_subagents`, `manage_task`, `multi_replace_file_content`,
+`read_url_content`, `replace_file_content`, `run_command`, `schedule`,
+`search_web`, `send_message`, `view_file`, `write_to_file`.
+
+What this changes vs the old Gemini CLI picture:
+
+- **agy has real subagents** (`invoke_subagent`/`define_subagent`/
+  `manage_subagents` + `send_message`) and background-task management
+  (`manage_task`). → `fanout` demoted further; probably never needed.
+- **agy has `schedule`** — one-shot timer or cron expression that wakes
+  the agent with a prompt (same idea as Claude's `ScheduleWakeup`).
+  Confirmed limits, from its schema: it cannot run commands itself,
+  it is **in-memory and dies with the session**, and it cannot reach
+  the phone. → `cronr` (persistent systemd timers) and `notifyr` are
+  still needed; `schedule` complements them within a session.
+- **agy has `generate_image`** — a *reverse* gap: Claude Code has no
+  native image generation. Nothing to build; just worth knowing.
+- No MCP-resource tools, no memory tool and no glob in its toolset
+  (grep/list_dir cover finding files).
+
+Behavior tests run in a scratch arena:
+
+| Test | Result |
+|---|---|
+| Enumerate tools | Clean list of 19 (above) |
+| Edit a text file | Worked, auto-approved in trusted folder |
+| Edit a Jupyter cell, keep `.ipynb` valid | **Passed** — it wrote a `python3 -c` json script rather than text-replacing. Notebook stayed valid. Cost: a per-command approval each time → `nbcell` demoted to nice-to-have (stable prefix + no ad-hoc python). |
+| Wait for a file to appear | Worked via a hand-rolled `for … sleep 1` shell loop — a unique command string needing fresh approval. → exactly the `waitfor` case. |
+| Asked agy which CLI tools *it* wants for the homelab | Its list: ntfy client, `tea` (Gitea CLI), `skopeo`/`crane` (registry), `ofelia`/cron daemon, `ctop`-style fleet status — near-1:1 with section 4, and it independently made the approval-prefix argument. |
+
+**Buy before build:** agy's suggestions overlap with off-the-shelf
+tools. Evaluate first: `tea` (official Gitea CLI — but it does not read
+the Pi5's zst action logs and has no `wait-quiet`, which stay
+`giteactl`'s reason to exist, possibly as a thin layer *on top of*
+`tea`), `skopeo`/`crane` (cover most of `reghelper` — remaining value
+is prune *plans* and size summaries), `ctop` (interactive TUI, not
+agent-friendly output — `fleet` still wins for agents).
+
+## 3. Gaps → tools to build
+
+Ordered by expected value. Each becomes its own folder + binary +
+README, per repo convention.
+
+### 3.1 `cronr` — scheduled/recurring runs *(Claude: `CronCreate`/`CronList`/`CronDelete`, `/loop`)*
+
+agy's built-in `schedule` dies with the session (see section 2).
+`cronr` manages **systemd user timers** so an agent can create
+recurring or one-shot jobs that survive session exit and reboot
+(including "run this prompt every morning" via `agy -p …`).
+
+```
+cronr add nightly-ci-check --schedule "*-*-* 07:00" --cmd 'agy -p "check CI status and notify"'
+cronr add once-reboot-check --at "2026-08-06 03:00" --cmd '…'   # one-shot
+cronr list                    # name, schedule, next run, last result
+cronr logs nightly-ci-check   # journalctl for the unit
+cronr rm nightly-ci-check
+```
+
+Output prints the generated unit files so the result is verifiable.
+No daemon of its own — systemd does the running.
+
+### 3.2 `waitfor` — block until a condition holds *(Claude: `Monitor`)*
+
+Turns "poll every N seconds" into a single blocking tool call, so the
+agent doesn't burn turns polling.
+
+```
+waitfor --cmd "curl -sf https://gitea.brasse-pc.eu/api/healthz" --interval 30s --timeout 20m
+waitfor --cmd "ssh pi5 docker ps --format '{{.Names}}'" --matches 'gitea' --timeout 10m
+waitfor … --then 'notifyr send --msg "gitea is back up"'
+```
+
+Exit 0 = condition met, exit 3 = timeout; last output is printed either
+way. `--then` runs a command on success (composes with `notifyr`).
+
+### 3.3 `notifyr` — push notifications, send **and read** *(Claude: `PushNotification`)*
+
+The ntfy server **already runs on the Pi5** and is the house-wide
+notification bus (topics like `Info`, `pi5-server-fel`, `ci-fel` — see
+infra-Doc `services/observability.md`). `notifyr` is a thin client so
+every agent uses it the same way:
+
+```
+notifyr send --topic ci-fel --title "Build failed" --msg "agent-tools arm64 test: FAIL" --priority high
+notifyr read --topic pi5-server-fel --since 2h     # poll mode: what has alerted lately?
+```
+
+`read` (ntfy's `?poll=1&since=…`) is the underrated half: it lets an
+agent *check what the infra has been complaining about* before/after a
+change. Config (`~/.config/notifyr/config.json`): server URL + token.
+
+### 3.4 `pagepub` — publish an HTML/Markdown report to a URL *(Claude: `Artifact`)*
+
+Claude Code can publish reports as web pages; Gemini cannot. `pagepub`
+rsyncs a file to a static-file host on the Pi5 (nginx container behind
+NPM, e.g. `pages.brasse-pc.eu`) and prints the stable URL.
+
+```
+pagepub publish report.html --slug ci-report     → https://pages.brasse-pc.eu/ci-report/
+pagepub publish notes.md --slug pi5-audit        # .md rendered to HTML with built-in template
+pagepub list | rm <slug>
+```
+
+Requires the static host to exist first (small infra task; goes in
+infra-Doc + NPM proxy host via `npmctl`).
+
+### 3.5 `nbcell` — Jupyter notebook editing *(Claude: `NotebookEdit`)*
+
+`.ipynb` is JSON that's miserable to edit via `replace`. `nbcell`
+exposes cells as text:
+
+```
+nbcell list nb.ipynb                 # index, type, first line, exec count
+nbcell show nb.ipynb 3               # cell source (and outputs with --outputs)
+nbcell edit nb.ipynb 3 --from-file cell.py
+nbcell add nb.ipynb --at 4 --type code --from-file new.py
+nbcell rm nb.ipynb 7
+```
+
+### 3.6 `wtreectl` — disposable git worktrees *(Claude: worktree isolation for agents)*
+
+Claude Code can give each subagent an isolated git worktree. `wtreectl`
+does the same for any agent:
+
+```
+wtreectl new [--branch dev/foo]   # prints the new worktree path
+wtreectl list
+wtreectl clean                    # removes worktrees with no changes
+```
+
+Lets two agent sessions work in the same repo without trampling each
+other.
+
+### 3.7 `fanout` — parallel subagent orchestration *(Claude: `Workflow`, `Agent`)* — roadmap, not agreed
+
+Runs N prompts as parallel headless agent processes (`gemini -p` /
+`claude -p`) with a concurrency cap, collecting each result as JSON in
+an output dir. A poor man's `Workflow`:
+
+```
+fanout run jobs.json --max 3 --out results/
+```
+
+Heavier than the other tools and overlaps with agent-helm's territory —
+park until there's a concrete need.
+
+## 4. Suggested tools from infra history
+
+Grounded in what the agent has already been doing per infra-Doc
+(`maintenance-and-gaps.md`, `services/source-control-and-deploy.md`,
+`services/observability.md`, the per-service "operational quick-ref"
+blocks). These help **any** agent (Claude or Gemini) administer the
+fleet, and they respect the read-only sudo policy
+(`ssh/claude-sudo-policy.md`): everything below is read-or-notify;
+mutations still go through Björn's supervised tmux flow.
+
+### 4.1 `giteactl` — Gitea repos, Actions runs and CI logs
+
+The biggest recurring friction. Today: CI status is polled ad hoc, and
+logs for private repos are only readable as zst files under
+`/srv/storage1/gitea/actions_log/…` on the Pi5. Wraps the Gitea REST +
+Actions API:
+
+```
+giteactl runs <repo> [--limit 5]          # status, branch, duration
+giteactl log <repo> <run> [--job N]       # fetches + decompresses the zst log
+giteactl wait <repo> [--timeout 30m]      # block until latest run finishes; exit 0 = green
+giteactl wait-quiet [--max-active 1]      # block until ≤N heavy builds are running
+giteactl release <repo> [<tag>]           # rolling-release assets + checksums
+```
+
+`wait-quiet` encodes the hard-learned rule "serialize pushes — >2 heavy
+builds take the Pi5 down": `giteactl wait-quiet && git push`.
+Composes with `waitfor`/`notifyr`.
+
+### 4.2 `fleet` — one-shot health snapshot of the whole homelab
+
+Every service doc ends with the same hand-rolled loop over
+`ssh pi5-claude sudo claude-docker ps/inspect …`. `fleet` does that
+loop once, properly:
+
+```
+fleet status          # containers (state, image, restarts), disk/mergerfs fill %, failed systemd units
+fleet status --host brasse-linux01
+fleet checks          # Uptime Kuma monitor states + last ntfy alerts (via notifyr read)
+```
+
+Read-only by construction (claude-docker wrapper + sudo allowlist), so
+it needs no new permissions. Output is a stable text table an agent can
+diff between runs.
+
+### 4.3 `envaudit` — compose ↔ `.env` key auditor
+
+Grounded in a real incident (`${STORAGE1}` undefined in `/srv/.env` →
+bad mount, rollback) and in maintenance-and-gaps' inline-secret
+findings:
+
+```
+envaudit check /srv/dockge-staks --env /srv/.env
+  → UNDEFINED  ${STORAGE1}   used by media-stack/compose.yaml
+  → INLINE     LDAP_ADMIN_PASSWORD  hardcoded in openldap/compose.yaml (should live in .env)
+  → UNUSED     OLD_API_KEY   defined but referenced nowhere
+```
+
+Pure text analysis of compose files — safe to run anywhere, catches the
+two failure classes that have actually happened.
+
+### 4.4 `reghelper` — docker-registry catalog & hygiene
+
+The LAN registry (`192.168.0.19:5000`) has no UI, no auth and no
+cleanup story, and the backup plan explicitly wants it slimmed:
+
+```
+reghelper ls                      # catalog + tags + image sizes
+reghelper tags <image>
+reghelper prune-plan --keep 2     # prints the delete+GC commands (does NOT run them)
+```
+
+`prune-plan` deliberately only *prints* the mutation commands for the
+supervised tmux flow — same pattern as the sudo policy.
+
+### 4.5 Backup status reader — once Backrest/restic exists
+
+future-plans.md has the whole Backrest+restic design chosen but
+unbuilt. When it lands, a `fleet backups` subcommand (last snapshot age
+per source, repo size, last check result) closes the loop — an agent
+can then *verify* backups instead of trusting them. Not a separate
+tool; park under `fleet`.
+
+### Cross-reference
+
+`/home/brasse/repos/dify-agent-tools/` already has a specced-but-unbuilt
+set of FastAPI tools (file/image sorting, face recognition, ST-card
+export…) for the Dify platform. Different runtime (HTTP tools vs CLI
+binaries), same philosophy — don't duplicate those here.
+
+## 5. Suggested build order
+
+1. `notifyr` — smallest, everything else composes with it, ntfy already runs.
+2. `giteactl` — removes the biggest daily friction (CI logs + build serialization).
+3. `waitfor` + `cronr` — turns both agents into unattended operators.
+4. `fleet` — replaces the hand-rolled health loops in every runbook.
+5. `envaudit`, `reghelper`, `pagepub`, `nbcell`, `wtreectl` — as needed.
+6. `fanout` — only if a concrete multi-agent need shows up.

notifyr/README.md

@@ -0,0 +1,62 @@
+# notifyr — ntfy client for agents
+
+Send and **read** notifications on the homelab's ntfy bus
+(`ntfy.brasse-pc.eu`) with one stable command prefix, so any agent can
+alert a human — and check what the infrastructure has been complaining
+about — without hand-rolled `curl` loops that each need a fresh
+approval. Closes the `PushNotification` gap from
+[`doc/tool-parity.md`](../doc/tool-parity.md) §3.3.
+
+## Usage
+
+```
+notifyr send --msg "text" [--topic T] [--title X]
+             [--priority min|low|default|high|urgent] [--tags a,b]
+notifyr read [--topic T] [--since 10m|2h|all] [--limit N]
+notifyr topics          # known topics + which one is the default
+notifyr version
+```
+
+Examples:
+
+```bash
+notifyr send --topic ci-fel --title "Build failed" \
+  --msg "agent-tools arm64 test: FAIL" --priority high
+
+notifyr read --topic pi5-server-fel --since 2h   # what has alerted lately?
+notifyr read --limit 5                           # newest 5 on the default topic
+```
+
+`read` uses ntfy's poll mode (`/json?poll=1&since=…`) and prints one
+greppable line per message:
+
+```
+2026-08-07 00:29:40 [high] (Build failed) agent-tools arm64 test: FAIL  #warning
+```
+
+Exit codes: 0 ok, 1 error (bad config, server unreachable, invalid
+priority…). `read` with zero messages is **not** an error.
+
+## Config
+
+`~/.config/notifyr/config.json`, created with homelab defaults on the
+first run (`NOTIFYR_CONFIG` overrides the path):
+
+```json
+{
+  "server": "https://ntfy.brasse-pc.eu",
+  "token": "",
+  "default_topic": "claude",
+  "topics": { "ci-fel": "failed Gitea Actions builds", "…": "…" }
+}
+```
+
+`topics` is informational — it feeds `notifyr topics` so an agent can
+pick the right bus without reading infra-Doc first. Edit freely.
+
+## Build & test
+
+```bash
+go test ./...
+go build -o build/notifyr .
+```

notifyr/go.mod

@@ -0,0 +1,3 @@
+module gitea.brasse-pc.eu/brasse/agent-tools/notifyr
+
+go 1.24

notifyr/main.go

@@ -0,0 +1,132 @@
+// notifyr sends and reads notifications on the homelab's ntfy bus, so
+// any agent can alert a human and check recent infra alerts the same
+// way. See doc/tool-parity.md §3.3.
+package main
+
+import (
+	"flag"
+	"fmt"
+	"os"
+	"sort"
+	"strings"
+
+	"gitea.brasse-pc.eu/brasse/agent-tools/notifyr/notify"
+)
+
+var version = "dev"
+
+const usage = `notifyr - ntfy client for agents (send and read notifications)
+
+Usage:
+  notifyr send --msg "text" [--topic T] [--title X]
+               [--priority min|low|default|high|urgent] [--tags a,b]
+  notifyr read [--topic T] [--since 10m|2h|all] [--limit N]
+  notifyr topics                    list known topics (from the config)
+  notifyr version
+
+Config: ~/.config/notifyr/config.json (created on first run; override
+path with NOTIFYR_CONFIG). Holds server URL, optional token, the
+default topic and the known-topics table.
+
+Examples:
+  notifyr send --topic ci-fel --title "Build failed" --msg "agent-tools arm64: FAIL" --priority high
+  notifyr read --topic pi5-server-fel --since 2h     # what has alerted lately?
+`
+
+func main() {
+	if len(os.Args) < 2 {
+		fmt.Print(usage)
+		os.Exit(2)
+	}
+	cfgPath := notify.ConfigPath()
+	switch os.Args[1] {
+	case "send":
+		cmdSend(cfgPath, os.Args[2:])
+	case "read":
+		cmdRead(cfgPath, os.Args[2:])
+	case "topics":
+		cmdTopics(cfgPath)
+	case "version", "--version", "-v":
+		fmt.Println("notifyr", version)
+	case "help", "--help", "-h":
+		fmt.Print(usage)
+	default:
+		die("unknown command %q — run 'notifyr help'", os.Args[1])
+	}
+}
+
+func cmdSend(cfgPath string, args []string) {
+	fs := flag.NewFlagSet("send", flag.ExitOnError)
+	topic := fs.String("topic", "", "topic (default: default_topic from config)")
+	title := fs.String("title", "", "notification title")
+	msg := fs.String("msg", "", "message text (required)")
+	priority := fs.String("priority", "", "min|low|default|high|urgent or 1-5")
+	tags := fs.String("tags", "", "comma-separated tags/emoji shortcodes")
+	fs.Parse(args)
+	cfg, err := notify.LoadConfig(cfgPath)
+	if err != nil {
+		die("%v", err)
+	}
+	if *topic == "" {
+		*topic = cfg.DefaultTopic
+	}
+	var tagList []string
+	if *tags != "" {
+		tagList = strings.Split(*tags, ",")
+	}
+	if err := notify.New(cfg.Server, cfg.Token).Send(*topic, *title, *msg, *priority, tagList); err != nil {
+		die("%v", err)
+	}
+	fmt.Printf("sent to %s/%s\n", cfg.Server, *topic)
+}
+
+func cmdRead(cfgPath string, args []string) {
+	fs := flag.NewFlagSet("read", flag.ExitOnError)
+	topic := fs.String("topic", "", "topic (default: default_topic from config)")
+	since := fs.String("since", "12h", "how far back: 10m, 2h, unix timestamp or all")
+	limit := fs.Int("limit", 0, "print at most N (newest) messages")
+	fs.Parse(args)
+	cfg, err := notify.LoadConfig(cfgPath)
+	if err != nil {
+		die("%v", err)
+	}
+	if *topic == "" {
+		*topic = cfg.DefaultTopic
+	}
+	msgs, err := notify.New(cfg.Server, cfg.Token).Read(*topic, *since, *limit)
+	if err != nil {
+		die("%v", err)
+	}
+	if len(msgs) == 0 {
+		fmt.Printf("no messages on %s since %s\n", *topic, *since)
+		return
+	}
+	for _, m := range msgs {
+		fmt.Println(notify.Format(m))
+	}
+}
+
+func cmdTopics(cfgPath string) {
+	cfg, err := notify.LoadConfig(cfgPath)
+	if err != nil {
+		die("%v", err)
+	}
+	names := make([]string, 0, len(cfg.Topics))
+	for n := range cfg.Topics {
+		names = append(names, n)
+	}
+	sort.Strings(names)
+	for _, n := range names {
+		mark := "  "
+		if n == cfg.DefaultTopic {
+			mark = "* "
+		}
+		fmt.Printf("%s%-18s %s\n", mark, n, cfg.Topics[n])
+	}
+	fmt.Fprintln(os.Stderr, "(* = default topic)")
+}
+
+func die(format string, args ...interface{}) {
+	fmt.Fprintf(os.Stderr, "notifyr: "+format+"\n", args...)
+	os.Exit(1)
+}

notifyr/notify/config.go

@@ -0,0 +1,78 @@
+package notify
+
+import (
+	"encoding/json"
+	"fmt"
+	"os"
+	"path/filepath"
+)
+
+// Config is ~/.config/notifyr/config.json, created with homelab
+// defaults on first run. NOTIFYR_CONFIG overrides the path.
+type Config struct {
+	Server       string            `json:"server"`
+	Token        string            `json:"token"`
+	DefaultTopic string            `json:"default_topic"`
+	Topics       map[string]string `json:"topics"` // known topics -> what they carry (informational)
+}
+
+func DefaultConfig() *Config {
+	return &Config{
+		Server:       "https://ntfy.brasse-pc.eu",
+		Token:        "",
+		DefaultTopic: "claude",
+		Topics: map[string]string{
+			"claude":           "agents' direct notes to Björn",
+			"agent-helm":       "agent-helm events (question waiting, session died)",
+			"Info":             "*arr system events",
+			"media-hamtningar": "media grabbed for download",
+			"media-nytt":       "new media landed in Jellyfin",
+			"pi5-server":       "server maintenance (reboots, watchtower)",
+			"pi5-server-fel":   "server problems: failed units, disk space",
+			"ci-fel":           "failed Gitea Actions builds",
+			"monitoring":       "Uptime Kuma up/down alerts",
+		},
+	}
+}
+
+func ConfigPath() string {
+	if p := os.Getenv("NOTIFYR_CONFIG"); p != "" {
+		return p
+	}
+	dir, err := os.UserConfigDir()
+	if err != nil {
+		dir = "."
+	}
+	return filepath.Join(dir, "notifyr", "config.json")
+}
+
+// LoadConfig reads the config, creating it with defaults on first run.
+func LoadConfig(path string) (*Config, error) {
+	cfg := DefaultConfig()
+	data, err := os.ReadFile(path)
+	if os.IsNotExist(err) {
+		if err := SaveConfig(path, cfg); err != nil {
+			return nil, err
+		}
+		fmt.Fprintf(os.Stderr, "notifyr: created %s\n", path)
+		return cfg, nil
+	}
+	if err != nil {
+		return nil, err
+	}
+	if err := json.Unmarshal(data, cfg); err != nil {
+		return nil, fmt.Errorf("%s: %w", path, err)
+	}
+	return cfg, nil
+}
+
+func SaveConfig(path string, cfg *Config) error {
+	if err := os.MkdirAll(filepath.Dir(path), 0o755); err != nil {
+		return err
+	}
+	data, err := json.MarshalIndent(cfg, "", "  ")
+	if err != nil {
+		return err
+	}
+	return os.WriteFile(path, append(data, '\n'), 0o600)
+}

notifyr/notify/notify.go

@@ -0,0 +1,165 @@
+// Package notify is a thin client for a ntfy server: publish
+// notifications and poll past ones, so agents can both alert humans
+// and check what the infrastructure has been complaining about.
+package notify
+
+import (
+	"bufio"
+	"encoding/json"
+	"fmt"
+	"io"
+	"net/http"
+	"net/url"
+	"strings"
+	"time"
+)
+
+// Message is one ntfy message as returned by the /json poll endpoint.
+type Message struct {
+	ID       string   `json:"id"`
+	Time     int64    `json:"time"`
+	Event    string   `json:"event"`
+	Topic    string   `json:"topic"`
+	Title    string   `json:"title"`
+	Message  string   `json:"message"`
+	Priority int      `json:"priority"`
+	Tags     []string `json:"tags"`
+}
+
+// Client talks to one ntfy server.
+type Client struct {
+	Server string // e.g. https://ntfy.brasse-pc.eu
+	Token  string // optional bearer token
+	HTTP   *http.Client
+}
+
+func New(server, token string) *Client {
+	return &Client{
+		Server: strings.TrimRight(server, "/"),
+		Token:  token,
+		HTTP:   &http.Client{Timeout: 30 * time.Second},
+	}
+}
+
+func (c *Client) auth(req *http.Request) {
+	if c.Token != "" {
+		req.Header.Set("Authorization", "Bearer "+c.Token)
+	}
+}
+
+// ValidPriority reports whether p is a priority ntfy accepts.
+func ValidPriority(p string) bool {
+	switch p {
+	case "", "1", "2", "3", "4", "5", "min", "low", "default", "high", "max", "urgent":
+		return true
+	}
+	return false
+}
+
+// Send publishes a message to a topic.
+func (c *Client) Send(topic, title, msg, priority string, tags []string) error {
+	if topic == "" {
+		return fmt.Errorf("no topic given (flag --topic or default_topic in the config)")
+	}
+	if msg == "" {
+		return fmt.Errorf("empty message")
+	}
+	if !ValidPriority(priority) {
+		return fmt.Errorf("invalid priority %q (use min|low|default|high|urgent or 1-5)", priority)
+	}
+	req, err := http.NewRequest("POST", c.Server+"/"+url.PathEscape(topic), strings.NewReader(msg))
+	if err != nil {
+		return err
+	}
+	c.auth(req)
+	if title != "" {
+		req.Header.Set("Title", title)
+	}
+	if priority != "" {
+		req.Header.Set("Priority", priority)
+	}
+	if len(tags) > 0 {
+		req.Header.Set("Tags", strings.Join(tags, ","))
+	}
+	resp, err := c.HTTP.Do(req)
+	if err != nil {
+		return err
+	}
+	defer resp.Body.Close()
+	body, _ := io.ReadAll(io.LimitReader(resp.Body, 4096))
+	if resp.StatusCode >= 300 {
+		return fmt.Errorf("server answered %s: %s", resp.Status, strings.TrimSpace(string(body)))
+	}
+	return nil
+}
+
+// Read polls past messages from a topic. since accepts ntfy's formats:
+// a duration ("10m", "2h"), a unix timestamp, a message id, or "all".
+func (c *Client) Read(topic, since string, limit int) ([]Message, error) {
+	if topic == "" {
+		return nil, fmt.Errorf("no topic given (flag --topic or default_topic in the config)")
+	}
+	if since == "" {
+		since = "all"
+	}
+	u := fmt.Sprintf("%s/%s/json?poll=1&since=%s", c.Server, url.PathEscape(topic), url.QueryEscape(since))
+	req, err := http.NewRequest("GET", u, nil)
+	if err != nil {
+		return nil, err
+	}
+	c.auth(req)
+	resp, err := c.HTTP.Do(req)
+	if err != nil {
+		return nil, err
+	}
+	defer resp.Body.Close()
+	if resp.StatusCode >= 300 {
+		body, _ := io.ReadAll(io.LimitReader(resp.Body, 4096))
+		return nil, fmt.Errorf("server answered %s: %s", resp.Status, strings.TrimSpace(string(body)))
+	}
+	var out []Message
+	sc := bufio.NewScanner(resp.Body)
+	sc.Buffer(make([]byte, 0, 64*1024), 1024*1024)
+	for sc.Scan() {
+		line := strings.TrimSpace(sc.Text())
+		if line == "" {
+			continue
+		}
+		var m Message
+		if err := json.Unmarshal([]byte(line), &m); err != nil {
+			continue // tolerate junk lines; poll output is one JSON object per line
+		}
+		if m.Event != "message" {
+			continue
+		}
+		out = append(out, m)
+	}
+	if err := sc.Err(); err != nil {
+		return nil, err
+	}
+	if limit > 0 && len(out) > limit {
+		out = out[len(out)-limit:] // keep the newest
+	}
+	return out, nil
+}
+
+// Format renders a message as one stable, greppable line.
+func Format(m Message) string {
+	ts := time.Unix(m.Time, 0).Format("2006-01-02 15:04:05")
+	prio := ""
+	switch {
+	case m.Priority >= 4:
+		prio = " [high]"
+	case m.Priority > 0 && m.Priority <= 2:
+		prio = " [low]"
+	}
+	title := ""
+	if m.Title != "" {
+		title = " (" + m.Title + ")"
+	}
+	tags := ""
+	if len(m.Tags) > 0 {
+		tags = "  #" + strings.Join(m.Tags, " #")
+	}
+	return fmt.Sprintf("%s%s%s %s%s", ts, prio, title, strings.ReplaceAll(m.Message, "\n", " ⏎ "), tags)
+}

notifyr/notify/notify_test.go

@@ -0,0 +1,135 @@
+package notify
+
+import (
+	"fmt"
+	"net/http"
+	"net/http/httptest"
+	"path/filepath"
+	"strings"
+	"testing"
+)
+
+func TestSendSetsHeadersAndBody(t *testing.T) {
+	var got *http.Request
+	var body string
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		got = r
+		b := make([]byte, 1024)
+		n, _ := r.Body.Read(b)
+		body = string(b[:n])
+		fmt.Fprint(w, `{"id":"x"}`)
+	}))
+	defer srv.Close()
+
+	c := New(srv.URL, "tok123")
+	err := c.Send("ci-fel", "Build failed", "arm64 test: FAIL", "high", []string{"warning", "ci"})
+	if err != nil {
+		t.Fatal(err)
+	}
+	if got.URL.Path != "/ci-fel" {
+		t.Errorf("path = %q", got.URL.Path)
+	}
+	if body != "arm64 test: FAIL" {
+		t.Errorf("body = %q", body)
+	}
+	for hdr, want := range map[string]string{
+		"Title":         "Build failed",
+		"Priority":      "high",
+		"Tags":          "warning,ci",
+		"Authorization": "Bearer tok123",
+	} {
+		if v := got.Header.Get(hdr); v != want {
+			t.Errorf("%s = %q, want %q", hdr, v, want)
+		}
+	}
+}
+
+func TestSendValidation(t *testing.T) {
+	c := New("http://example.invalid", "")
+	if err := c.Send("", "", "hello", "", nil); err == nil {
+		t.Error("empty topic accepted")
+	}
+	if err := c.Send("t", "", "", "", nil); err == nil {
+		t.Error("empty message accepted")
+	}
+	if err := c.Send("t", "", "hello", "banana", nil); err == nil {
+		t.Error("bogus priority accepted")
+	}
+}
+
+func TestReadParsesPollOutput(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		if r.URL.Query().Get("poll") != "1" {
+			t.Errorf("poll param missing: %s", r.URL.RawQuery)
+		}
+		if r.URL.Query().Get("since") != "2h" {
+			t.Errorf("since = %q", r.URL.Query().Get("since"))
+		}
+		fmt.Fprintln(w, `{"id":"a","time":1754500000,"event":"message","topic":"t","message":"first","priority":3}`)
+		fmt.Fprintln(w, `{"id":"b","time":1754500060,"event":"keepalive","topic":"t"}`)
+		fmt.Fprintln(w, `not json at all`)
+		fmt.Fprintln(w, `{"id":"c","time":1754500120,"event":"message","topic":"t","title":"T","message":"second","priority":5,"tags":["x"]}`)
+	}))
+	defer srv.Close()
+
+	msgs, err := New(srv.URL, "").Read("t", "2h", 0)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(msgs) != 2 {
+		t.Fatalf("got %d messages, want 2 (keepalive + junk filtered)", len(msgs))
+	}
+	if msgs[0].Message != "first" || msgs[1].Title != "T" {
+		t.Errorf("unexpected messages: %+v", msgs)
+	}
+}
+
+func TestReadLimitKeepsNewest(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		for i := 1; i <= 5; i++ {
+			fmt.Fprintf(w, "{\"id\":\"%d\",\"time\":%d,\"event\":\"message\",\"topic\":\"t\",\"message\":\"m%d\"}\n", i, 1754500000+i, i)
+		}
+	}))
+	defer srv.Close()
+	msgs, err := New(srv.URL, "").Read("t", "all", 2)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(msgs) != 2 || msgs[0].Message != "m4" || msgs[1].Message != "m5" {
+		t.Errorf("limit should keep the newest: %+v", msgs)
+	}
+}
+
+func TestFormatIsOneGreppableLine(t *testing.T) {
+	line := Format(Message{Time: 1754500000, Title: "Backup", Message: "done\nall good", Priority: 4, Tags: []string{"ok"}})
+	if strings.Contains(line, "\n") {
+		t.Error("format must be a single line")
+	}
+	for _, want := range []string{"[high]", "(Backup)", "done", "#ok"} {
+		if !strings.Contains(line, want) {
+			t.Errorf("line %q missing %q", line, want)
+		}
+	}
+}
+
+func TestConfigRoundtrip(t *testing.T) {
+	path := filepath.Join(t.TempDir(), "config.json")
+	cfg, err := LoadConfig(path) // first run creates defaults
+	if err != nil {
+		t.Fatal(err)
+	}
+	if cfg.Server == "" || cfg.DefaultTopic == "" {
+		t.Error("defaults incomplete")
+	}
+	cfg.DefaultTopic = "elsewhere"
+	if err := SaveConfig(path, cfg); err != nil {
+		t.Fatal(err)
+	}
+	again, err := LoadConfig(path)
+	if err != nil {
+		t.Fatal(err)
+	}
+	if again.DefaultTopic != "elsewhere" {
+		t.Error("saved change did not persist")
+	}
+}