Review SPARTA conversation transcripts and grading with full transparency. Use when asked to review conversations, show transcripts, compare student vs teacher grades, inspect persona evaluations, or export sessions for human annotation. Replaces the black-box opacity of raw JSONL session files with rich turn-by-turn rendering including entity gate decisions, QRA citations, self-grade iterations, and persona evaluation reasoning.
Standardmäßig ist der Prompt ausgewählt, der zuerst die Quelle prüft. Sie können zu einem direkten Befehl wechseln oder eine lokale Kopie herunterladen.
Quelldateien prüfen
Lesen Sie SKILL.md und alle von SkillsMP angezeigten Begleitdateien, bevor Sie sich für eine Installation entscheiden.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Ein direkter Befehl überspringt den Prüf-Prompt. Prüfen Sie die Quelle, bevor Sie ihn ausführen.
Review SPARTA conversation transcripts and grading with full transparency. Use when asked to review conversations, show transcripts, compare student vs teacher grades, inspect persona evaluations, or export sessions for human annotation. Replaces the black-box opacity of raw JSONL session files with rich turn-by-turn rendering including entity gate decisions, QRA citations, self-grade iterations, and persona evaluation reasoning.
allowed-tools
["Bash","Read"]
triggers
["review conversation","show conversation transcript","compare student teacher grades","inspect persona evaluation","export conversation markdown","conversation disagreements","review sparta session","find conversation where","show me a conversation","which conversations"]
metadata
{"short-description":"Transparent conversation transcript viewer with grading","version":"3.0.0"}
Transparent conversation transcript viewer for SPARTA stress test sessions.
Every turn, grade, and evaluation visible — no black boxes.
Quick Start
# List available session files
./run.sh files
# Show summary of latest session file
./run.sh list
# Show full transcript of session #3
./run.sh show -n 3
# Show specific session by ID
./run.sh show -i stress_20260222_175146
# Show only disagreements (student vs teacher)
./run.sh compare
# Export to markdown for human annotation
./run.sh export -o review.md
# NDJSON stream for automation
./run.sh json-stream
Commands
Command
Description
files
List available session JSONL files with sizes and counts
list [file]
Summary table: all sessions with grades, agreement, resolution
show [file]
Full transcript with turn-by-turn metadata and grading
compare [file]
Filter to only student-teacher disagreements
export [file]
Export sessions as annotated markdown
json-stream [file]
NDJSON output with merged shadow/delta data
find [file]
Structured search with filters + free-text (agent-human collaboration)
search <query>
Semantic search via /memory recall (BM25 + embedding + multi-hop)
ingest [file]
Learn session summaries to /memory for semantic search
brief [file]
Plain markdown for in-chat display (no Rich/ANSI)
flow [file]
Mermaid conversation flow diagrams with inline grading metadata
data [file]
Export JSON for /create-figure (radar, heatmap, bar charts)
Evaluation reasoning (why the persona accepted or rejected)
For system turns (SPARTA answering):
Full response text with entity gate decisions
Self-grade: letter grade + composite score + iteration count
QRA citation count
Issues flagged during self-grading
Rationale from semantic grader
What You See Per Session
Session grade: 8-dimension rubric breakdown with visual bars
Student vs Teacher comparison: side-by-side grades, confidence, model names
Agreement signal: AGREE (green) or DISAGREE (red)
Improvement delta: first → final score, outer/inner iteration counts
QRA citation metrics: verified/total, grounding average
Data Sources
File
Location
Content
sessions_*.jsonl
sparta-stress-test/results/sessions/
Full conversation transcripts
shadow.jsonl
~/.pi/assistant/
Teacher grades + agreement
shadow_deltas.jsonl
~/.pi/assistant/
Improvement tracking per session
Environment Variables
Variable
Default
Description
SESSIONS_DIR
../sparta-stress-test/results/sessions/
Sessions directory
SHADOW_JSONL_PATH
~/.pi/assistant/shadow.jsonl
Shadow entries
SHADOW_DELTA_PATH
~/.pi/assistant/shadow_deltas.jsonl
Improvement deltas
Semantic Search via /memory (search, ingest)
Composes with /memory for BM25 + semantic embedding + multi-hop graph traversal.
Far superior to bespoke text search for natural language queries.
Workflow
# 1. After a stress test, ingest sessions to /memory
./run.sh ingest
# 2. Search semantically (natural language)
./run.sh search "conversations where Brandon missed a fake control"
./run.sh search "GPS spoofing questions that got resolved"
./run.sh search "low-scoring sessions with iteration improvement"# 3. Or use --semantic on find (combines structured filters + semantic)
./run.sh find --semantic "fake control missed" --grade F
How It Works
Layer
What It Does
BM25
Keyword matching against ingested session summaries
Semantic embedding
Dense vector similarity for meaning-based recall
Multi-hop traversal
Graph edges connecting related sessions, controls, personas
ingest extracts problem/solution pairs from each session and stores them
with tags (grade_A, difficulty_complex, resolution_resolved, etc.)
for filtered recall.
Graceful Degradation
If /memory is unavailable (no ArangoDB, no embedding service), semantic
commands return empty results and fall back to structured filters. The skill
never crashes — it just loses the semantic layer.
Agent-Human Collaboration (find)
The find command is the primary collaboration surface. When the human asks
a question about conversations, the agent translates it into structured filters:
# "Show me a conversation where Brandon didn't flag a non-existent control"
./run.sh find --eval flaw_missed --resolution no_coverage
# "Show me where Margaret failed the first response but iteration saved it to 90%"
./run.sh find --first-eval "!satisfactory" --min-composite 0.9
# "Which complex questions got fully resolved?"
./run.sh find --difficulty complex --resolution resolved
# "Find conversations about SV-SP-1"
./run.sh find --text "SV-SP-1"# "Show disagreements on flawed questions"
./run.sh find --difficulty flawed --disagree
# "How many F-grade sessions are there?"
./run.sh find --grade F --format count
# "Show low-scoring sessions in Rich detail"
./run.sh find --max-composite 0.5 --format show
Find Filters
Flag
Description
-g, --grade A|B|C|F
Filter by session grade
-r, --resolution
resolved, partial, no_coverage, ambiguous
-e, --eval
Persona evaluation: satisfactory, incomplete, wrong, flaw_caught, flaw_missed. Prefix ! to negate
--first-eval
First persona evaluation only (supports ! negation)
--min-composite
Minimum final composite score (0.0-1.0)
--max-composite
Maximum final composite score (0.0-1.0)
-d, --difficulty
simple, medium, complex, flawed, ambiguous
-p, --persona
Filter by persona name
--min-turns
Minimum turn count (higher = more iteration)
-t, --text
Free-text search across all turn content, evaluations, rationales
-s, --semantic
Semantic search via /memory recall (BM25 + embedding + multi-hop). Use for natural language queries
--agree/--disagree
Filter by teacher agreement
-f, --format
Output: brief (markdown), show (Rich), count (just count)
All filters are AND-combined. Default output is brief (pasteable markdown).
Translation Guide for Agents
When the human asks in natural language, translate to filters:
Export JSON files that /create-figure consumes directly:
# Export all chart data
./run.sh data -o /tmp/charts/
# Then visualize with /create-figurecd .pi/skills/create-figure
./run.sh radar --input /tmp/charts/radar.json --output radar.pdf
./run.sh heatmap --input /tmp/charts/heatmap.json --output heatmap.pdf
./run.sh metrics --input /tmp/charts/grade_distribution.json --type bar
# Export only radar chart data
./run.sh data -c radar -o /tmp/charts/
JSON files produced:
radar.json — 8-dimension rubric scores per session (for radar charts)
heatmap.json — difficulty x grade distribution (for heatmaps)
grade_distribution.json — grade letter counts (for bar charts)
