History src/analyze.js · TreeTrace

Code Commits Tags

History: src/analyze.js

← back to file · 24 commits

Strengthen analysis-layer signal detection accuracy

Zion Boggan · Jun 20, 2026 (2 days ago)

2dacbee

Release 0.10.0: --each batch export, --deterministic, broader refusal capture

Zion Boggan · Jun 19, 2026 (3 days ago)

76d0f78

Address v0.9.0 audit findings: schema export, detection recall, redaction, docs

Zion Boggan · Jun 19, 2026 (3 days ago)

cde6644

feat: capture typed rejections and refusals

Zion Boggan · Jun 18, 2026 (4 days ago)

d80486a

feat: shared schema version, structured exit codes, MCP tree tool

Zion Boggan · Jun 18, 2026 (4 days ago)

3d1ebdb

refactor: decompose SECURITY_INTENT and RISKY_CMD regexes into named pieces

Zion Boggan · Jun 18, 2026 (4 days ago)

a7d34c9

perf: drop analyzeTree and buildTree from O(N^2) to O(N)

Zion Boggan · Jun 18, 2026 (4 days ago)

411615f

Artifact diet: trim verbose rendering across report, security, handoff, prompt-tree, lessons, and agent-memory

Zion Boggan · Jun 15, 2026 (1 week ago)

c1a4d74

Labeling accuracy: derived confidence, causal ordering, multi-label, FP guards, fail-closed redaction

Zion Boggan · Jun 15, 2026 (1 week ago)

81d5699

Harden redaction and fix independent audit findings (0.5.0)

Zion Boggan · Jun 13, 2026 (1 week ago)

535e0f8

Add --security report, deterministic hallucination detector, and read-only MCP server

Zion Boggan · Jun 13, 2026 (1 week ago)

3548f10

Release 0.4.1: fix redaction hex-secret leak and tighten security-signal precision

Zion Boggan · Jun 13, 2026 (1 week ago)

17a9eda

Filter the handoff decisions list to real strategic directions

Zion Boggan · Jun 13, 2026 (1 week ago)

c4cb33e

Synthesize concrete, correctly-attributed lesson bodies

Zion Boggan · Jun 13, 2026 (1 week ago)

7b477a5

Detect destructive-then-recovery bad paths and pick a real next-work direction

Zion Boggan · Jun 13, 2026 (1 week ago)

a96c1d8

Extract real user-enforced constraints for agent memory

Zion Boggan · Jun 13, 2026 (1 week ago)

7281781

Tie failure signals to corroboration and time-valid semantic linkage

Zion Boggan · Jun 13, 2026 (1 week ago)

e2e52a2

Detect security regressions from action content and prompt intent

Zion Boggan · Jun 13, 2026 (1 week ago)

1a981ec

Analysis v2: surface captured reasoning-block count in the report

Zion Boggan · Jun 12, 2026 (1 week ago)

32a2d36

Analysis v2: verified security audit trail from agent actions, model attribution, tiered confidence

Zion Boggan · Jun 12, 2026 (1 week ago)

b61cfed

Escape untrusted content in Markdown output (prevent HTML injection)

Zion Boggan · Jun 12, 2026 (1 week ago)

d1df111

Derive agent memory from the session instead of hardcoding project notes

Zion Boggan · Jun 12, 2026 (1 week ago)

ebdfdd8

Fix analysis heuristics and document Node requirement

Zion Boggan · Jun 12, 2026 (1 week ago)

a655d37

Add analysis layer: failure signals, correction chains, lessons, eval candidates, agent memory

Zion Boggan · Jun 12, 2026 (1 week ago)

9a4a23e