v0.8.0

0.8.0 (2026-08-28)

Added

• bench: P3.7 walled scorecard with honest per-suite pass rates (#111) (b13ca30)
• bench: TTY Rich scorecard with per-case stderr ticks (#113) (7517c4b)
• cli: doctor progress ticks (#103) (c28daab)
• cli: P4.6 phase ticks for transcribe/grab/get (#109) (adabfc8)
• cli: P4.6 search progress ticks (#110) (3c36dca)
• cli: progress ticks on read + mcp serve banner (#100) (5ab2c81)
• install: write user-scope reach rule for Claude and Cursor (#86) (76ce119)
• mcp: MCPServer on mcp 2.x (P2.3) (#98) (afbe15a)
• mcp: tell agents to retry walled hits via read_url (#88) (1c03d0e)

Fixed

• bench: ladder ticks during a long case (#114) (e80c0c3)
• benchmark: thin content is not a pass; label smoke suite (#79) (9dc6af4)
• channels: drop routing theater; doctor is a probe (#85) (e5e32b1)
• config: drop dead knobs; load .env without override (#81) (3b4f333)
• config: environment variables beat YAML (#82) (02463b9)
• deps: pin curl_cffi, trafilatura, and ddgs (#83) (c365dbc)
• doctor: honest stealth probe; stop skill install side effect (#74) (102a94c)
• mcp: SSRF guards on MCP URL tools (#105) (545ea35)
• release: fail if RELEASE_PLEASE_TOKEN is missing (#84) (0fdeed0)
• skill: keep SKILL.md description under 1024 chars (#89) (403ac8d)
• unlocker: domain memory TTL and unpin on failure (#102) (f565ee0)
• unlocker: drop hardcoded Chrome 126 UA (#104) (08515f0)
• unlocker: fail loud on thin and challenge pages (#93) (89fbb5e)
• unlocker: login shells are not a pass (#116) (fcac810)
• unlocker: P3.5 Jina opt-out (default on) (#107) (2013a4b)
• unlocker: Playwright off asyncio for MCP (P3.10) (#96) (2809ac7)

What's Changed

• test: make must-not-be-called tripwires able to fail by @capad-xyz in #73
• fix(doctor): honest stealth probe; stop skill install side effect by @capad-xyz in #74
• fix(benchmark): thin content is not a pass; label smoke suite by @capad-xyz in #79
• docs: add frozen honesty plan by @capad-xyz in #80
• fix(config): drop dead knobs; load .env without override by @capad-xyz in #81
• fix(config): environment variables beat YAML by @capad-xyz in #82
• fix(deps): pin curl_cffi, trafilatura, and ddgs by @capad-xyz in #83
• fix(release): fail if RELEASE_PLEASE_TOKEN is missing by @capad-xyz in #84
• fix(channels): drop routing theater; doctor is a probe by @capad-xyz in #85
• feat(install): write user-scope reach rule for Claude and Cursor by @capad-xyz in #86
• ci: slim PR checks; full matrix on main by @capad-xyz in #87
• feat(mcp): tell agents to retry walled hits via read_url by @capad-xyz in #88
• fix(skill): keep SKILL.md description under 1024 chars by @capad-xyz in #89
• docs(plan): tick P1.1 P1.2 P1.2b; note PyPI 0.7.3 by @capad-xyz in #90
• docs(plan): pipx keep / uvx try by @capad-xyz in #91
• fix(unlocker): fail loud on thin and challenge pages by @capad-xyz in #93
• refactor(mcp): FastMCP on mcp 1.x (P2.1) by @capad-xyz in #94
• docs(plan): F9 localhost HTTP later by @capad-xyz in #92
• docs(plan): eval + P3.10 + X2/X3 status by @capad-xyz in #95
• fix(unlocker): Playwright off asyncio for MCP (P3.10) by @capad-xyz in #96
• ci: mcp 2.x compat job (P2.2) by @capad-xyz in #97
• feat(mcp): MCPServer on mcp 2.x (P2.3) by @capad-xyz in #98
• feat(cli): progress ticks on read + mcp serve banner by @capad-xyz in #100
• docs(plan): P4.6 CLI UX feedback (doctor/search/transcribe/grab) by @capad-xyz in #101
• fix(unlocker): domain memory TTL and unpin on failure by @capad-xyz in #102
• fix(unlocker): drop hardcoded Chrome 126 UA by @capad-xyz in #104
• feat(cli): doctor progress ticks by @capad-xyz in #103
• fix(mcp): SSRF guards on MCP URL tools by @capad-xyz in #105
• docs(plan): park P3.6b redirect/DNS-rebinding SSRF by @capad-xyz in #106
• fix(unlocker): P3.5 Jina opt-out (default on) by @capad-xyz in #107
• docs(plan): mark X3 posted; refresh P1.3 / X2 notes by @capad-xyz in #108
• feat(p4.6): search progress ticks for CLI UX by @capad-xyz in #110
• feat(cli): P4.6 phase ticks for transcribe/grab/get by @capad-xyz in #109
• feat(bench): P3.7 walled scorecard with honest per-suite pass rates by @capad-xyz in #111
• docs(plan): P4.6 applies to new long runners by @capad-xyz in #112
• feat(bench): TTY Rich scorecard with per-case stderr ticks by @capad-xyz in #113
• fix(bench): ladder ticks during a long case by @capad-xyz in #114
• docs: F8 pipx keep / uvx try by @capad-xyz in #115
• fix(unlocker): login shells are not a pass by @capad-xyz in #116
• docs(plan): park P3.11 stealth page.content nav race by @capad-xyz in #117
• docs(plan): park F11 Windows MCP shim vs editable install by @capad-xyz in https://gi...

v0.7.2

0.7.2 (2026-08-04)

Fixed

• share: capture full Grok share ids and read Perplexity Pages (#71) (7106ed1)

What's Changed

• ci(release): add a manual dispatch path so automated tags can still publish by @capad-xyz in #69
• chore: bump python-dotenv from 1.2.1 to 1.2.2 by @dependabot[bot] in #65
• chore: bump requests from 2.32.5 to 2.34.2 by @dependabot[bot] in #63
• chore: bump yt-dlp from 2025.5.22 to 2026.7.4 by @dependabot[bot] in #64
• ci: bump googleapis/release-please-action from 4 to 5 by @dependabot[bot] in #62
• docs: sync the --human and ChatGPT share docs with 0.7.1 behaviour by @capad-xyz in #70
• fix(share): capture full Grok share ids and read Perplexity Pages by @capad-xyz in #71
• chore(main): release 0.7.2 by @github-actions[bot] in #72

Full Changelog: v0.7.1...v0.7.2

v0.7.1

0.7.1 (2026-08-04)

Fixed

• mcp: cap the SDK below 2.0, which breaks the server on startup (#60) (607db0a)
• unlocker: read ChatGPT /s/ shares and stop soft walls swallowing --human (#68) (9057628)

What's Changed

• ci: halve CI runs, cache pip, scope token, add dependabot by @capad-xyz in #49
• ci(release): automate release prep with release-please by @capad-xyz in #56
• ci(release): only cut a release when the package changes by @capad-xyz in #58
• ci: bump softprops/action-gh-release from 2 to 3 by @dependabot[bot] in #50
• ci: bump actions/checkout from 4 to 7 by @dependabot[bot] in #51
• chore: bump mypy from 1.19.1 to 2.3.0 by @dependabot[bot] in #53
• chore: bump rich from 14.3.2 to 15.0.0 by @dependabot[bot] in #54
• chore: bump types-pyyaml from 6.0.12.20250915 to 6.0.12.20260724 by @dependabot[bot] in #55
• ci: bump actions/setup-python from 5 to 7 by @dependabot[bot] in #52
• docs: sync contributor and agent docs with the CI and release setup by @capad-xyz in #59
• fix(mcp): cap the SDK below 2.0, which breaks the server on startup by @capad-xyz in #60
• docs: link the full demo video below the README hero by @capad-xyz in #66
• fix(unlocker): read ChatGPT /s/ shares and stop soft walls swallowing --human by @capad-xyz in #68
• chore(main): release 0.7.1 by @github-actions[bot] in #61

New Contributors

• @dependabot[bot] made their first contribution in #50
• @github-actions[bot] made their first contribution in #61

Full Changelog: v0.7.0...v0.7.1

v0.7.0

What's Changed

• docs(changelog): credit #21 and #23 in the 0.6.0 notes by @capad-xyz in #29
• ci(release): republish server.json to the MCP registry on tag push by @capad-xyz in #34
• test(unlocker): add CDN challenge phrase matrix by @terminalchai in #35
• docs(benchmark): explain scorecard results by @terminalchai in #36
• feat(unlocker): carry response headers by @terminalchai in #39
• feat(unlocker): detect Cloudflare challenges by @terminalchai in #40
• feat(unlocker): extract complete conversations from AI-chat share links by @capad-xyz in #41
• docs(readme): use pepy.tech downloads badge by @capad-xyz in #42
• docs(contributing): add block phrase guide by @terminalchai in #43
• ci: enforce clean ruff checks by @terminalchai in #45
• fix(cli): exit cleanly on Ctrl+C and broken pipe by @capad-xyz in #46
• fix(types): clear mypy findings and gate on it in CI by @capad-xyz in #47
• release: 0.7.0 by @capad-xyz in #48

Full Changelog: v0.6.0...v0.7.0

v0.6.0

What's Changed

• feat(benchmark): show category pass rates by @terminalchai in #23
• feat: doctor reports whether searchts is wired into local AI agents by @capad-xyz in #21
• feat(read): add source-receipt fields fetched_at and final_url by @tapheret2 in #25
• feat: add Fastly/Akamai/CF challenge block phrases by @tapheret2 in #26
• test(benchmark): add python-docs and httpbin cases by @tapheret2 in #27
• chore: release v0.6.0 (source receipts + challenge phrases + benchmark cases) by @capad-xyz in #28

New Contributors

• @tapheret2 made their first contribution in #25

Full Changelog: v0.5.2...v0.6.0

v0.5.2

What's Changed

• docs: absolute URLs in README so images/links render on PyPI by @capad-xyz in #19
• chore: release v0.5.2 (MCP registry metadata + PyPI README fix) by @capad-xyz in #20

Full Changelog: v0.5.1...v0.5.2

v0.5.1

What's Changed

• fix: suggest close CLI subcommands by @terminalchai in #16
• chore: release v0.5.1 (CLI suggestions + benchmark scorecard) by @capad-xyz in #18

New Contributors

• @terminalchai made their first contribution in #16

Full Changelog: v0.5.0...v0.5.1

v0.5.0

What's Changed

• chore: capad contact identity + README hero upgrade by @capad-xyz in #2
• docs: plain bullets for Why searchts by @capad-xyz in #3
• docs: hero demo GIF by @capad-xyz in #4
• docs(contributing): add 'what we merge / what we don't' section by @capad-xyz in #5
• docs: update tagline to "The missing layer between AI and the web" by @capad-xyz in #6
• feat: make the MCP/agent surface legible (tool descriptions, get_status, docs) by @capad-xyz in #7
• chore: claim + containerize the MCP server for Glama by @capad-xyz in #8
• feat: contribution valve (issue/PR templates, ROADMAP, version-sync CI) by @capad-xyz in #9
• feat: unlocker benchmark (proof + regression canary) by @capad-xyz in #10
• chore: release v0.5.0 by @capad-xyz in #11

Full Changelog: v0.4.1...v0.5.0

v0.4.1

Full Changelog: v0.4.0...v0.4.1

v0.4.0

Full Changelog: v0.3.1...v0.4.0