personal memory agent
health: document sol call health pipeline + reframe description
apps/health/call.py exposes `sol call health pipeline` (the only
command actually defined by the health app itself) but the skill
never documented it. Add a pipeline-summary section with --day /
--yesterday flags and examples.
Reframe description from "diagnose" infrastructure framing to task
framing ("monitor service uptime, troubleshoot capture gaps"). Make
the tri-scope surface explicit (sol health*, sol talent*,
sol call health*) since health diagnostics cross levels. Add a
Gotchas block (callosum 10s timeout, ms-precision talent log IDs,
today-vs-yesterday pipeline window). Expand TRIGGER.
Part of skills audit req_loq3e2lk pass 1+2.
+41
-9
+41
-9
apps/health/talent/health/SKILL.md
+41
-9
apps/health/talent/health/SKILL.md
···
1
1
---
2
2
name: health
3
3
description: >
4
-
Diagnose solstone service health, inspect agent run logs, and check system
5
-
status. View service uptimes, crashes, queue depths, recent errors, and
6
-
agent run costs. Includes a journal layout reference for navigating data
7
-
files. Use when the owner reports issues, asks about service health, agent
8
-
costs, pipeline status, or when troubleshooting capture gaps and processing
9
-
failures.
4
+
Monitor service uptime, troubleshoot capture gaps and processing failures,
5
+
review agent run costs and errors, and summarize think-pipeline health
6
+
for a day. Covers three CLI surfaces: `sol health` (service supervisor
7
+
status + logs), `sol talent` (agent run logs + details), and
8
+
`sol call health pipeline` (per-day pipeline summary). Includes a journal
9
+
layout reference for navigating data files.
10
10
TRIGGER: health, status, is it running, something broke, service down,
11
-
errors, agent runs, costs, logs, pipeline, diagnostics, system check.
11
+
uptime, errors, agent runs, costs, logs, pipeline, pipeline anomalies,
12
+
diagnostics, system check, troubleshoot capture gap, sol health,
13
+
sol health logs, sol talent logs, sol talent log,
14
+
sol call health pipeline.
12
15
---
13
16
14
17
# Health CLI Skill
15
18
16
-
Use these commands to check service health, view logs, and inspect agent runs.
19
+
Monitor solstone service uptime, troubleshoot failures, and inspect agent runs. Invoke via Bash: `sol health ...`, `sol talent ...`, or `sol call health <command>`.
20
+
21
+
**Scope note**: Three CLI surfaces live here: `sol health*` (supervisor/service level), `sol talent*` (agent run level), and `sol call health <command>` (app-level pipeline health). They're grouped together because health troubleshooting routinely crosses the three levels.
17
22
18
-
**Typical workflow**: `sol health` to check service status → `sol health logs` to inspect recent log output → `sol talent logs` to review agent runs → `sol talent log <ID>` for run details.
23
+
**Typical workflow**: `sol health` → `sol health logs` → `sol talent logs` → `sol talent log <ID>` for agent-run detail → `sol call health pipeline` for a day-level pipeline summary.
19
24
20
25
## status
21
26
···
115
120
sol talent log 1700000000001 --full
116
121
```
117
122
123
+
## pipeline summary
124
+
125
+
```bash
126
+
sol call health pipeline [--day YYYYMMDD | --yesterday]
127
+
```
128
+
129
+
Summarize think-pipeline health for one day — anomalies, performance metrics, and per-stage outcomes across the day's processing runs. Emits JSON.
130
+
131
+
- `--day YYYYMMDD`: target day. Defaults to today.
132
+
- `--yesterday`: shortcut for yesterday. Mutually exclusive with `--day`.
133
+
134
+
Use this when you want a day-level view after daily processing completes, rather than a per-run drilldown via `sol talent log`.
135
+
136
+
Examples:
137
+
138
+
```bash
139
+
sol call health pipeline
140
+
sol call health pipeline --yesterday
141
+
sol call health pipeline --day 20260115
142
+
```
143
+
118
144
## journal layout
119
145
120
146
Reference map of key paths. `journal/` is the journal root.
···
183
209
184
210
### High agent costs
185
211
Run `sol talent logs --summary` for aggregated cost view. Filter by agent: `sol talent logs <agent-name> --summary`.
212
+
213
+
## Gotchas
214
+
215
+
- **`sol health` times out at 10 seconds.** If the supervisor is slow or hung, you'll hit the timeout before seeing results. Confirm the supervisor process is alive (`ps` / `sol supervisor` status) before assuming the service is down.
216
+
- **Talent log IDs are millisecond timestamps.** `sol talent log 1700000000001` expects the full ID from `sol talent logs`, not a seconds-precision value.
217
+
- **`sol call health pipeline` needs today's processing to have run.** Running it at 6am before the daily pipeline has executed will return sparse results for today; use `--yesterday` instead.