autoresearch

Name: autoresearch
Author: popmechanic

134stars6forks

How to install autoresearch?

popmechanic/vibesos/autoresearch

$npx -y skills add popmechanic/VibesOS --skill autoresearch

Installs into the current project.

Summary

Run the Parallel Autoresearch Engine — massively parallel SKILL.md optimization. Generates 10+ SKILL.md variants per generation, tests each with programmatic harness (no browser needed), scores with triple-run averaging, and iterates autonomously. Use when asked to run autoresear

Files · 1

View on GitHub

SKILL.md

1	# Parallel Autoresearch Engine
2
3	> Plan mode: This skill is ONE plan step: "Invoke /vibes:autoresearch". Do not decompose.
4
5	## Prerequisites Check
6
7	```bash
8	VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
9	echo "Checking autoresearch prerequisites..."
10	test -f "$VIBES_ROOT/scripts/eval-ssr-check.ts" && echo "✓ Tier 1.5 SSR check" \|\| echo "✗ Missing eval-ssr-check.ts"
11	test -f "$VIBES_ROOT/scripts/eval-harness.ts" && echo "✓ Tier 2 harness" \|\| echo "✗ Missing eval-harness.ts"
12	test -f "$VIBES_ROOT/scripts/eval-parallel.ts" && echo "✓ Orchestrator" \|\| echo "✗ Missing eval-parallel.ts"
13	test -f "$VIBES_ROOT/scripts/eval-scoring.ts" && echo "✓ Scoring" \|\| echo "✗ Missing eval-scoring.ts"
14	test -f "$VIBES_ROOT/eval/config.md" && echo "✓ Config" \|\| echo "✗ Missing eval/config.md"
15	test -f "$VIBES_ROOT/eval/napkin.md" && echo "✓ Napkin" \|\| echo "✗ Missing eval/napkin.md"
16	ls "$VIBES_ROOT/eval/specs/"*.md 2>/dev/null \| wc -l \| xargs -I{} echo "✓ {} eval specs found"
17	cd "$VIBES_ROOT/scripts" && bun -e "import React from 'react'; console.log('✓ React available')" 2>/dev/null \|\| echo "✗ React not installed"
18	```
19
20	If any prerequisite is missing, stop and inform the user.
21
22	## Running
23
24	### Option 1: Full Autonomous Run (Recommended)
25
26	Dispatch the autoresearch orchestrator agent (`.claude/agents/autoresearch-orchestrator.md`) with context from eval/config.md, eval/napkin.md, and eval/scoreboard.md.
27
28	Pass any CLI arguments from the user (e.g., `--variants=5 --generations=10`).
29
30	### Option 2: Single Eval Pipeline Test
31
32	```bash
33	VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
34	bun "$VIBES_ROOT/scripts/eval-parallel.ts" --mode=eval-only <app.jsx> <spec.md>
35	```
36
37	### Option 3: Score Existing Generation
38
39	```bash
40	VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
41	bun "$VIBES_ROOT/scripts/eval-scoring.ts" "$VIBES_ROOT/eval/results/gen-N/"
42	```
43
44	## What It Does
45
46	Each generation:
47	1. Mutate: N independent SKILL.md variants (fix-targeted, structural, adversarial, etc.)
48	2. Generate: Each variant × each prompt × 3 runs
49	3. Evaluate: Tier 1 (static) → Tier 1.5 (SSR) → Tier 2 (data model) — all programmatic
50	4. Score: Triple-run averaging with consistency penalty; fitness = mean - 0.5×stddev
51	5. Select: Best variant replaces current SKILL.md; git commit on improvement
52	6. Repeat: Until plateau (3 gens), max generations, or score oscillation
53
54	## Monitoring
55
56	- `eval/results/gen-N/summary.json` — per-generation results
57	- `eval/results/summaries.json` — cumulative history
58	- `eval/scoreboard.md` — human-readable scoreboard
59	- `eval/napkin.md` — failure log (grows monotonically)
60
61	## Final Report
62
63	```bash
64	VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
65	bun "$VIBES_ROOT/scripts/eval-report.ts" "$VIBES_ROOT/eval/results/summaries.json"
66	```

People who liked this also viewed

6 picks

Type

autoresearch

134stars6forks

View on GitHub

How to install autoresearch?

popmechanic/vibesos/autoresearch

$npx -y skills add popmechanic/VibesOS --skill autoresearch

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Parallel Autoresearch Engine
2
3	> Plan mode: This skill is ONE plan step: "Invoke /vibes:autoresearch". Do not decompose.
4
5	## Prerequisites Check
6
7	```bash
8	VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
9	echo "Checking autoresearch prerequisites..."
10	test -f "$VIBES_ROOT/scripts/eval-ssr-check.ts" && echo "✓ Tier 1.5 SSR check" \|\| echo "✗ Missing eval-ssr-check.ts"
11	test -f "$VIBES_ROOT/scripts/eval-harness.ts" && echo "✓ Tier 2 harness" \|\| echo "✗ Missing eval-harness.ts"
12	test -f "$VIBES_ROOT/scripts/eval-parallel.ts" && echo "✓ Orchestrator" \|\| echo "✗ Missing eval-parallel.ts"
13	test -f "$VIBES_ROOT/scripts/eval-scoring.ts" && echo "✓ Scoring" \|\| echo "✗ Missing eval-scoring.ts"
14	test -f "$VIBES_ROOT/eval/config.md" && echo "✓ Config" \|\| echo "✗ Missing eval/config.md"
15	test -f "$VIBES_ROOT/eval/napkin.md" && echo "✓ Napkin" \|\| echo "✗ Missing eval/napkin.md"
16	ls "$VIBES_ROOT/eval/specs/"*.md 2>/dev/null \| wc -l \| xargs -I{} echo "✓ {} eval specs found"
17	cd "$VIBES_ROOT/scripts" && bun -e "import React from 'react'; console.log('✓ React available')" 2>/dev/null \|\| echo "✗ React not installed"
18	```
19
20	If any prerequisite is missing, stop and inform the user.
21
22	## Running
23
24	### Option 1: Full Autonomous Run (Recommended)
25
26	Dispatch the autoresearch orchestrator agent (`.claude/agents/autoresearch-orchestrator.md`) with context from eval/config.md, eval/napkin.md, and eval/scoreboard.md.
27
28	Pass any CLI arguments from the user (e.g., `--variants=5 --generations=10`).
29
30	### Option 2: Single Eval Pipeline Test
31
32	```bash
33	VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
34	bun "$VIBES_ROOT/scripts/eval-parallel.ts" --mode=eval-only <app.jsx> <spec.md>
35	```
36
37	### Option 3: Score Existing Generation
38
39	```bash
40	VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
41	bun "$VIBES_ROOT/scripts/eval-scoring.ts" "$VIBES_ROOT/eval/results/gen-N/"
42	```
43
44	## What It Does
45
46	Each generation:
47	1. Mutate: N independent SKILL.md variants (fix-targeted, structural, adversarial, etc.)
48	2. Generate: Each variant × each prompt × 3 runs
49	3. Evaluate: Tier 1 (static) → Tier 1.5 (SSR) → Tier 2 (data model) — all programmatic
50	4. Score: Triple-run averaging with consistency penalty; fitness = mean - 0.5×stddev
51	5. Select: Best variant replaces current SKILL.md; git commit on improvement
52	6. Repeat: Until plateau (3 gens), max generations, or score oscillation
53
54	## Monitoring
55
56	- `eval/results/gen-N/summary.json` — per-generation results
57	- `eval/results/summaries.json` — cumulative history
58	- `eval/scoreboard.md` — human-readable scoreboard
59	- `eval/napkin.md` — failure log (grows monotonically)
60
61	## Final Report
62
63	```bash
64	VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
65	bun "$VIBES_ROOT/scripts/eval-report.ts" "$VIBES_ROOT/eval/results/summaries.json"
66	```

People who liked this also viewed

6 picks

Type