agent-research-aggregator

Name: agent-research-aggregator
Author: Ar9av

41installs86forks

How to install agent-research-aggregator?

ar9av/paperorchestra/agent-research-aggregator

$npx -y skills add Ar9av/PaperOrchestra --skill agent-research-aggregator

Installs into the current project.

Summary

Pre-pipeline aggregator that scans AI agent cache directories (.claude, .cursor, .antigravity, .openclaw) or any user-specified directory for experimentation logs, extracts insights and numeric results, and formats them as PaperOrchestra-ready inputs (idea.md + experimental_log.m

Files · 1

View on GitHub

SKILL.md

1	# agent-research-aggregator
2
3	---
4
5	## Should I run? (decision gate)
6
7	Before starting Phase 1, check whether aggregation is actually needed:
8
9	\| Situation \| Action \|
10	\|---\|---\|
11	\| `workspace/inputs/idea.md` and `workspace/inputs/experimental_log.md` both exist and are non-empty \| Skip this skill entirely. Proceed directly to `paper-orchestra`. \|
12	\| Either file is missing or empty, and the user provided a directory path \| Run this skill with that directory as `--search-roots`. \|
13	\| Either file is missing or empty, and no directory was provided \| Scan cwd and `~` by default; show the discovery summary to the user before continuing. \|
14	\| The inputs exist but look thin (e.g. idea.md has < 5 lines, no numeric data in experimental_log.md) \| Ask the user whether to supplement with aggregation or proceed as-is. \|
15
16	The skill is intentionally a pre-pass — it is cheap to skip and should only run when the structured inputs don't already exist.
17
18	---
19
20	A pre-processing skill for PaperOrchestra (arXiv:2604.05018). Reads scattered
21	experimentation artifacts from AI coding-agent cache directories and synthesizes
22	them into the structured `(I, E)` input pair the PaperOrchestra pipeline expects.
23
24	```
25	[.claude/] [.cursor/] [.antigravity/] [.openclaw/]
26	│ │ │ │
27	└────────────┴──────────────┴───────────────┘
28	│
29	Phase 1: Discovery
30	(discover_logs.py)
31	│
32	discovered_logs.json
33	│
34	Phase 2: Extraction
35	(LLM call per log batch)
36	│
37	raw_experiments.json
38	│
39	Phase 3: Synthesis
40	(LLM call — consolidate)
41	│
42	synthesis.json
43	│
44	Phase 4: Formatting
45	(format_po_inputs.py)
46	│
47	┌────────────┴────────────┐
48	workspace/inputs/ workspace/ara/
49	idea.md aggregation_report.md
50	experimental_log.md discovered_logs.json
51	raw_experiments.json
52	synthesis.json
53	```
54
55	The output drops directly into `workspace/inputs/` so the user can immediately
56	run `paper-orchestra` on the same workspace.
57
58	---
59
60	## Inputs
61
62	\| Parameter \| Required \| Default \| Description \|
63	\|---\|---\|---\|---\|
64	\| `--search-roots` \| no \| cwd, `~` \| Comma-separated directories to scan for agent caches \|
65	\| `--agents` \| no \| all \| Comma-separated subset: `claude,cursor,antigravity,openclaw` \|
66	\| `--workspace` \| no \| `./workspace` \| PaperOrchestra workspace root \|
67	\| `--depth` \| no \| 4 \| Max directory scan depth (prevents runaway scans on large home dirs) \|
68	\| `--since` \| no \| none \| Only include logs modified after this date (ISO 8601: `2025-01-01`) \|
69
70	The user specifies these when invoking the skill, or you may ask them for
71	`--search-roots` if the current directory has no detectable agent caches.
72
73	---
74
75	## Phase 1 — Discovery (deterministic)
76
77	Run the discovery script to catalog every relevant log file:
78
79	```bash
80	python skills/agent-research-aggregator/scripts/discover_logs.py \
81	--search-roots <roots> \
82	--agents <agents> \
83	--depth <depth> \
84	--since <since> \
85	--out workspace/ara/discovered_logs.json
86	```
87
88	The script exits with code 2 when no `--project` filter is set (this is
89	expected on the first run). It prints a "Projects found" list to stdout —
90	show it to the user immediately.
91
92	If no logs are found at all: stop and ask the user to specify
93	`--search-roots` or point you at a directory that contains agent cache folders.
94
95	---
96
97	## Phase 1.5 — Project Selection (mandatory)
98
99	**A paper can only be written from a single project. You must ask the user
100	which project to use before any LLM processing begins.**
101
102	1. Display the numbered project list from the discovery summary, e.g.:
103	```
104	Projects found:
105	[1] /home/alice/p

People who liked this also viewed

6 picks

Type

agent-research-aggregator

41installs86forks

View on GitHub

How to install agent-research-aggregator?

ar9av/paperorchestra/agent-research-aggregator

$npx -y skills add Ar9av/PaperOrchestra --skill agent-research-aggregator

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# agent-research-aggregator
2
3	---
4
5	## Should I run? (decision gate)
6
7	Before starting Phase 1, check whether aggregation is actually needed:
8
9	\| Situation \| Action \|
10	\|---\|---\|
11	\| `workspace/inputs/idea.md` and `workspace/inputs/experimental_log.md` both exist and are non-empty \| Skip this skill entirely. Proceed directly to `paper-orchestra`. \|
12	\| Either file is missing or empty, and the user provided a directory path \| Run this skill with that directory as `--search-roots`. \|
13	\| Either file is missing or empty, and no directory was provided \| Scan cwd and `~` by default; show the discovery summary to the user before continuing. \|
14	\| The inputs exist but look thin (e.g. idea.md has < 5 lines, no numeric data in experimental_log.md) \| Ask the user whether to supplement with aggregation or proceed as-is. \|
15
16	The skill is intentionally a pre-pass — it is cheap to skip and should only run when the structured inputs don't already exist.
17
18	---
19
20	A pre-processing skill for PaperOrchestra (arXiv:2604.05018). Reads scattered
21	experimentation artifacts from AI coding-agent cache directories and synthesizes
22	them into the structured `(I, E)` input pair the PaperOrchestra pipeline expects.
23
24	```
25	[.claude/] [.cursor/] [.antigravity/] [.openclaw/]
26	│ │ │ │
27	└────────────┴──────────────┴───────────────┘
28	│
29	Phase 1: Discovery
30	(discover_logs.py)
31	│
32	discovered_logs.json
33	│
34	Phase 2: Extraction
35	(LLM call per log batch)
36	│
37	raw_experiments.json
38	│
39	Phase 3: Synthesis
40	(LLM call — consolidate)
41	│
42	synthesis.json
43	│
44	Phase 4: Formatting
45	(format_po_inputs.py)
46	│
47	┌────────────┴────────────┐
48	workspace/inputs/ workspace/ara/
49	idea.md aggregation_report.md
50	experimental_log.md discovered_logs.json
51	raw_experiments.json
52	synthesis.json
53	```
54
55	The output drops directly into `workspace/inputs/` so the user can immediately
56	run `paper-orchestra` on the same workspace.
57
58	---
59
60	## Inputs
61
62	\| Parameter \| Required \| Default \| Description \|
63	\|---\|---\|---\|---\|
64	\| `--search-roots` \| no \| cwd, `~` \| Comma-separated directories to scan for agent caches \|
65	\| `--agents` \| no \| all \| Comma-separated subset: `claude,cursor,antigravity,openclaw` \|
66	\| `--workspace` \| no \| `./workspace` \| PaperOrchestra workspace root \|
67	\| `--depth` \| no \| 4 \| Max directory scan depth (prevents runaway scans on large home dirs) \|
68	\| `--since` \| no \| none \| Only include logs modified after this date (ISO 8601: `2025-01-01`) \|
69
70	The user specifies these when invoking the skill, or you may ask them for
71	`--search-roots` if the current directory has no detectable agent caches.
72
73	---
74
75	## Phase 1 — Discovery (deterministic)
76
77	Run the discovery script to catalog every relevant log file:
78
79	```bash
80	python skills/agent-research-aggregator/scripts/discover_logs.py \
81	--search-roots <roots> \
82	--agents <agents> \
83	--depth <depth> \
84	--since <since> \
85	--out workspace/ara/discovered_logs.json
86	```
87
88	The script exits with code 2 when no `--project` filter is set (this is
89	expected on the first run). It prints a "Projects found" list to stdout —
90	show it to the user immediately.
91
92	If no logs are found at all: stop and ask the user to specify
93	`--search-roots` or point you at a directory that contains agent cache folders.
94
95	---
96
97	## Phase 1.5 — Project Selection (mandatory)
98
99	**A paper can only be written from a single project. You must ask the user
100	which project to use before any LLM processing begins.**
101
102	1. Display the numbered project list from the discovery summary, e.g.:
103	```
104	Projects found:
105	[1] /home/alice/p

People who liked this also viewed

6 picks

Type