name	filtered-data
description	DEPRECATED — use data sub-agents with --pit directly. See DataSubAgents.md.
allowed-tools	Skill, Bash, Read
skills	neo4j-report, neo4j-xbrl, neo4j-news, neo4j-entity, neo4j-transcript, perplexity-search
context	fork
model	sonnet
user-invocable	false

DEPRECATED (2026-02-17) — Superseded by DataSubAgents architecture (--pit + pit_gate.py hooks). Use data sub-agents directly with --pit <ISO8601>. This skill is retained for reference only.

FILTER PROTOCOL (LEGACY)

Arguments: $ARGUMENTS

Step 1: PARSE

Extract: AGENT (after --agent), QUERY (after --query), PIT (from [PIT: datetime] if present)

Step 2: FETCH DATA

Call Skill tool: skill=AGENT, args=QUERY Store the COMPLETE response for validation.

Step 3: VALIDATE (MANDATORY)

YOU MUST EXECUTE THIS BASH COMMAND. DO NOT SKIP OR SIMULATE.

First, output "[VALIDATING]" on its own line.

Then create a temp file with the COMPLETE response and validate it:

cat > /tmp/validate_input.txt << 'DATAEOF'
<paste the COMPLETE response from Step 2 here - every line, unmodified>
DATAEOF

cat /tmp/validate_input.txt | /home/faisal/EventMarketDB/.claude/filters/validate.sh --source "AGENT_NAME" --pit "PIT_TIMESTAMP"

Example for perplexity-search:

cat > /tmp/validate_input.txt << 'DATAEOF'
[
  {"title": "Article 1", "url": "...", "snippet": "...", "date": "2023-02-20"},
  {"title": "Article 2", "url": "...", "snippet": "...", "date": "2023-02-21"}
]
DATAEOF

cat /tmp/validate_input.txt | /home/faisal/EventMarketDB/.claude/filters/validate.sh --source "perplexity-search" --pit "2023-02-22T16:52:26-05:00"

The script outputs:

"CLEAN" = data passes all checks
"CONTAMINATED:PIT_VIOLATION:YYYY-MM-DD" = date after PIT found (exit code 1)
"CONTAMINATED:FORBIDDEN_PATTERN:name" = forbidden field found (exit code 1)

CRITICAL: Show the exact output from the validation command. Do not paraphrase.

Step 4: RETURN

Check what the validation bash command output:

If the output was exactly "CLEAN":

Output "[VALIDATED:CLEAN]" on its own line
Return the complete data response from Step 2

If the output started with "CONTAMINATED:":

Output "[VALIDATED:CONTAMINATED]" on its own line
Extract the violation type (e.g., "PIT_VIOLATION" or "FORBIDDEN_PATTERN")
DO NOT return ANY data from the response
Return error: "Data validation failed: [violation_type] violated PIT constraint"
NEVER include specific dates, values, field contents, or article titles from contaminated data

Example of correct contamination handling:

Validation output: CONTAMINATED:PIT_VIOLATION:2025-06-04

Your response:
[VALIDATED:CONTAMINATED]
Data validation failed: PIT_VIOLATION violated PIT constraint

REDACTION RULE: When contamination detected, report ONLY the violation type. NEVER mention specific dates, values, article titles, URLs, or any content from the contaminated data.

Execute steps 1-4 in order. Show [VALIDATING] and [VALIDATED:*] markers.

Más de este repositorio

mismo repositorio

earnings-orchestrator

faisalanjum/EventTrader

Predict stock direction post 8-K earnings & refine using 10-Q/10-K outcomes

2026-05-083

earnings-prediction

faisalanjum/EventTrader

Predict stock direction after an 8-K earnings release from a prebuilt earnings context bundle

2026-05-043

earnings-learner

faisalanjum/EventTrader

Post-event causal attribution — explains why a stock moved after 8-K earnings, compares against prediction, writes reusable lessons for future predictions. Production invocation via SDK embed (main session), not fork.

2026-05-043

neo4j-schema

faisalanjum/EventTrader

Core Neo4j schema reference with all labels, relationships, data types, and indexes. Use when exploring database structure, checking field types, or understanding the financial knowledge graph schema.

2026-04-243

test-v1105-long-desc

faisalanjum/EventTrader

V1105_LONGDESC_START_MARKER. This skill tests v2.1.105's skill-description cap raise from 250 to 1536 characters. If this description appears truncated in the skill listing system-reminder at around character 250, the cap was not raised; if it appears truncated at around 1536, the cap was raised correctly per changelog. The test sentinel markers are placed at specific character offsets: V1105_OFFSET_100 is at char 100 approximately, V1105_OFFSET_300 is at char 300 approximately (old cap would truncate this), V1105_OFFSET_700 is at char 700 approximately, V1105_OFFSET_1000 is at char 1000, V1105_OFFSET_1200 is at char 1200, V1105_OFFSET_1400 is at char 1400, and V1105_END_MARKER is the last visible sentinel. This description also contains deliberately redundant filler to reach the target length without substantive content since the purpose is only to verify truncation behavior rather than to provide meaningful guidance. filler filler filler filler filler filler filler filler filler filler filler filler filler

2026-04-193

test-v1107-nest-child

faisalanjum/EventTrader

Child skill for v2.1.107 nesting retest — writes marker and returns

2026-04-193

name	filtered-data
description	DEPRECATED — use data sub-agents with --pit directly. See DataSubAgents.md.
allowed-tools	Skill, Bash, Read
skills	neo4j-report, neo4j-xbrl, neo4j-news, neo4j-entity, neo4j-transcript, perplexity-search
context	fork
model	sonnet
user-invocable	false

DEPRECATED (2026-02-17) — Superseded by DataSubAgents architecture (--pit + pit_gate.py hooks). Use data sub-agents directly with --pit <ISO8601>. This skill is retained for reference only.

FILTER PROTOCOL (LEGACY)

Arguments: $ARGUMENTS

Step 1: PARSE

Extract: AGENT (after --agent), QUERY (after --query), PIT (from [PIT: datetime] if present)

Step 2: FETCH DATA

Call Skill tool: skill=AGENT, args=QUERY Store the COMPLETE response for validation.

Step 3: VALIDATE (MANDATORY)

YOU MUST EXECUTE THIS BASH COMMAND. DO NOT SKIP OR SIMULATE.

First, output "[VALIDATING]" on its own line.

Then create a temp file with the COMPLETE response and validate it:

cat > /tmp/validate_input.txt << 'DATAEOF'
<paste the COMPLETE response from Step 2 here - every line, unmodified>
DATAEOF

cat /tmp/validate_input.txt | /home/faisal/EventMarketDB/.claude/filters/validate.sh --source "AGENT_NAME" --pit "PIT_TIMESTAMP"

Example for perplexity-search:

cat > /tmp/validate_input.txt << 'DATAEOF'
[
  {"title": "Article 1", "url": "...", "snippet": "...", "date": "2023-02-20"},
  {"title": "Article 2", "url": "...", "snippet": "...", "date": "2023-02-21"}
]
DATAEOF

cat /tmp/validate_input.txt | /home/faisal/EventMarketDB/.claude/filters/validate.sh --source "perplexity-search" --pit "2023-02-22T16:52:26-05:00"

The script outputs:

"CLEAN" = data passes all checks
"CONTAMINATED:PIT_VIOLATION:YYYY-MM-DD" = date after PIT found (exit code 1)
"CONTAMINATED:FORBIDDEN_PATTERN:name" = forbidden field found (exit code 1)

CRITICAL: Show the exact output from the validation command. Do not paraphrase.

Step 4: RETURN

Check what the validation bash command output:

If the output was exactly "CLEAN":

Output "[VALIDATED:CLEAN]" on its own line
Return the complete data response from Step 2

If the output started with "CONTAMINATED:":

Output "[VALIDATED:CONTAMINATED]" on its own line
Extract the violation type (e.g., "PIT_VIOLATION" or "FORBIDDEN_PATTERN")
DO NOT return ANY data from the response
Return error: "Data validation failed: [violation_type] violated PIT constraint"
NEVER include specific dates, values, field contents, or article titles from contaminated data

Example of correct contamination handling:

Validation output: CONTAMINATED:PIT_VIOLATION:2025-06-04

Your response:
[VALIDATED:CONTAMINATED]
Data validation failed: PIT_VIOLATION violated PIT constraint

REDACTION RULE: When contamination detected, report ONLY the violation type. NEVER mention specific dates, values, article titles, URLs, or any content from the contaminated data.

Execute steps 1-4 in order. Show [VALIDATING] and [VALIDATED:*] markers.