personal memory agent
feat: add --segments flag to transcripts read for span queries
Wires up cluster_span() to the CLI so agents can read multiple segments
in one call. Updates TODO Detector to use --segments $activity_segments
instead of per-segment iteration, completing the sources removal.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
+30
-16
+5
-9
apps/todos/muse/todo.md
+5
-9
apps/todos/muse/todo.md
···
38
38
- `sol call todos upcoming` – view upcoming todos to avoid duplicates
39
39
40
40
### Transcript Commands
41
-
- `sol call transcripts read --segment SEGMENT_KEY --transcripts` – read audio transcript for a segment
42
-
- `sol call transcripts read --segment SEGMENT_KEY --agents` – read agent outputs (screen summaries) for a segment
43
-
- `sol call transcripts read --segment SEGMENT_KEY --full` – read everything for a segment
41
+
- `sol call transcripts read --segments $activity_segments --transcripts` – read audio transcripts for this activity
42
+
- `sol call transcripts read --segments $activity_segments --agents` – read agent outputs (screen summaries) for this activity
43
+
- `sol call transcripts read --segment SEGMENT_KEY --full` – read everything for a single segment
44
44
- `sol call journal search QUERY` – cross-reference journal content
45
45
46
-
Activity segments: $activity_segments
47
-
48
46
**Query syntax**: Terms are AND'd by default; use `OR` for alternatives, quote phrases for exact matches, append `*` for prefix matching.
49
47
50
48
## Process
···
53
51
54
52
Read the activity's transcript and current todo state before making any changes.
55
53
56
-
1. For each segment in this activity ($activity_segments), load the transcript:
57
-
`sol call transcripts read --segment SEGMENT_KEY --transcripts`
58
-
Also load screen agent output for context:
59
-
`sol call transcripts read --segment SEGMENT_KEY --agents`
54
+
1. Load the activity's transcript and screen agent context:
55
+
`sol call transcripts read --segments $activity_segments --transcripts --agents --max 0`
60
56
2. Call `sol call todos list` to see the current checklist
61
57
3. Call `sol call todos upcoming -l 50` to check what's already scheduled
62
58
+18
-2
apps/transcripts/call.py
+18
-2
apps/transcripts/call.py
···
17
17
cluster_range,
18
18
cluster_scan,
19
19
cluster_segments,
20
+
cluster_span,
20
21
)
21
22
from think.utils import (
22
23
day_dirs,
···
85
86
segment: str | None = typer.Option(
86
87
None, "--segment", help="Segment key (HHMMSS_LEN, default: SOL_SEGMENT env)."
87
88
),
89
+
segments: str | None = typer.Option(
90
+
None, "--segments", help="Comma-separated segment keys for a span."
91
+
),
88
92
stream: str | None = typer.Option(
89
93
None, "--stream", help="Stream name (default: SOL_STREAM env)."
90
94
),
···
140
144
else:
141
145
sources = {"transcripts": True, "percepts": False, "agents": True}
142
146
143
-
if segment and (start or length is not None):
144
-
typer.echo("Error: Cannot mix --segment with --start/--length.", err=True)
147
+
# Validate mutually exclusive selection modes
148
+
mode_count = sum([
149
+
segment is not None,
150
+
segments is not None,
151
+
start is not None or length is not None,
152
+
])
153
+
if mode_count > 1:
154
+
typer.echo(
155
+
"Error: Cannot mix --segment, --segments, and --start/--length.",
156
+
err=True,
157
+
)
145
158
raise typer.Exit(1)
146
159
147
160
if (start is not None) != (length is not None):
···
154
167
start_dt = datetime.strptime(start, "%H%M%S")
155
168
end_dt = start_dt + timedelta(minutes=length)
156
169
markdown = cluster_range(day, start, end_dt.strftime("%H%M%S"), sources)
170
+
elif segments is not None:
171
+
span = [s.strip() for s in segments.split(",") if s.strip()]
172
+
markdown, _counts = cluster_span(day, span, sources, stream=stream)
157
173
elif segment is not None:
158
174
markdown, _counts = cluster_period(day, segment, sources, stream=stream)
159
175
else:
+7
-5
apps/transcripts/muse/transcripts/SKILL.md
+7
-5
apps/transcripts/muse/transcripts/SKILL.md
···
68
68
## read
69
69
70
70
```bash
71
-
sol call transcripts read [DAY] [--start HHMMSS --length MINUTES] [--segment KEY] [--stream NAME] [--full] [--raw] [--audio] [--screen] [--agents]
71
+
sol call transcripts read [DAY] [--start HHMMSS --length MINUTES] [--segment KEY] [--segments KEYS] [--stream NAME] [--full] [--raw] [--audio] [--screen] [--agents]
72
72
```
73
73
74
-
Read transcript content for a day, time range, or segment.
74
+
Read transcript content for a day, time range, segment, or span.
75
75
76
76
- `DAY`: day in `YYYYMMDD` (default: `SOL_DAY` env).
77
77
78
78
Read modes (mutually exclusive):
79
79
80
80
1. `--start HHMMSS --length MINUTES`: range mode. Both flags are required together; end time is auto-computed.
81
-
2. `--segment KEY`: segment mode using key from `segments`.
82
-
3. No time flags: whole-day mode.
81
+
2. `--segment KEY`: single segment mode using key from `segments`.
82
+
3. `--segments KEY1,KEY2,...`: span mode — comma-separated segment keys, merged in time order.
83
+
4. No time flags: whole-day mode.
83
84
84
85
Source flags:
85
86
···
92
93
93
94
Rules:
94
95
95
-
- Do not combine `--segment` with `--start/--length`.
96
+
- Do not combine `--segment`, `--segments`, or `--start/--length` with each other.
96
97
- Do not combine `--full` or `--raw` with individual source flags.
97
98
98
99
Source type meanings:
···
107
108
sol call transcripts read 20260115
108
109
sol call transcripts read 20260115 --start 090000 --length 30 --raw
109
110
sol call transcripts read 20260115 --segment 091500_300 --full
111
+
sol call transcripts read 20260115 --segments 091500_300,092000_300,092500_300 --transcripts --agents
110
112
sol call transcripts read 20260115 --audio
111
113
```
112
114