diff --git a/.release-please-manifest.json b/.release-please-manifest.json
index 76d5538a..b28fea99 100644
--- a/.release-please-manifest.json
+++ b/.release-please-manifest.json
@@ -1,3 +1,3 @@
 {
-  ".": "0.9.0"
+  ".": "0.9.1"
 }
diff --git a/CHANGELOG.md b/CHANGELOG.md
new file mode 100644
index 00000000..f7bf276b
--- /dev/null
+++ b/CHANGELOG.md
@@ -0,0 +1,168 @@
+# Changelog
+
+## [0.9.1](https://github.com/KeyValueSoftwareSystems/agent-opfor/compare/agent-opfor-v0.9.0...agent-opfor-v0.9.1) (2026-06-26)
+
+
+### Features
+
+* Add attackerInstructions config field for user-guided attack generation and judgment ([fd3bef1](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/fd3bef16420c500533507d8f1f88f8222e28008c))
+* add chat input detection and interaction functionality ([0edc335](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/0edc33561def426bf2ef7ee66378dd274426fb37))
+* Add custom HTTP header support to agent mode redteaming ([#35](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/35)) ([b32b2e7](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/b32b2e7dbe56b7f52dc54d2e81847539eb0ab547))
+* add DeepSeek and Azure OpenAI provider support ([a88d692](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/a88d692e8f915fa1abf9c2b7ff1549b93aae295c))
+* add drop down to extension for searching models ([9833b21](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/9833b21487367a1d556dfe6226eb27371c4131e8))
+* add ERROR verdict, multi-turn adaptive attacks, and overhauled HTML report ([48e687e](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/48e687e37de11247b0db6c1d79f6fc7a2d0970c1))
+* add evaluators for second-order content injection and server-side request forgery ([31cf557](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/31cf557f7ce377c03297b74915133f6ef67691a6))
+* add GitHub Actions CI workflows ([#77](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/77)) ([5c5f005](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/5c5f0054a97693ea0e6f075a93f393873402222e))
+* add initial project structure with core documentation, configuration examples, and licensing information ([84e84a8](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/84e84a80cc84ea42e4d1d3bc640145dfc4c95050))
+* Add live UI dashboard for autonomous runner ([9a81e32](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/9a81e32a7f6827487b28c1b6cdd380c50e2c3682))
+* add MCP target to SVG ([1bf5f87](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/1bf5f8792d3a9a3feb46ee20c51aa82efb744bd8))
+* Add MITRE ATLAS tagging, ID validation, coverage reporting, and evaluator updates across agent and MCP suites. ([#56](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/56)) ([be7121c](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/be7121c9e6fc6eff1f25d3aeafc089312c5f2c11))
+* add Netra telemetry connector with provider adapter pattern for setup trace curation ([fcfbb8f](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/fcfbb8f13913353eda10eaca68fca83025a2ee1d))
+* add Netra telemetry connector with provider adapter pattern for setup trace curation ([2baf30c](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/2baf30c71a6c09d55bdede806f53b3a2060a4e34))
+* add owasp suites and evaluators ([bf3818a](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/bf3818a7bcea38be39a4862cdd815dd2d5d7bb6e))
+* add owasp-api, output-trust, eu-ai-act-bias, agent-side mcp suites — 32 new evaluators ([f216d6e](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/f216d6ebc78f4a0ab492c81e04c49b1f0cb81bc3))
+* add per-attack judgeHint generated by attacker LLM for smarter, context-aware judge evaluation ([9007108](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/90071082a2a9a96932b453a300d16efc767eca14))
+* add precommit validator for evaluators ([3182217](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/3182217ca94d760d650cb70cf79c9940aa51b3b7))
+* add static packs to skills for code analysis ([d3c309b](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/d3c309bb11632ffe7629447b6df984b5938c00c0))
+* add static packs to skills for code analysis ([f8d4abb](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/f8d4abba4e326355fa6cbe2fb7dc39b541b732be))
+* add support for mcp tool filtering ([73bc73f](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/73bc73fcc0b12449e3aa4cfc9b82aa98b48ebcac))
+* add support for traces to MCP server ([cdca0d5](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/cdca0d5137ab50448735a637195307247613d675))
+* add support for traces to MCP server ([9c871ea](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/9c871ea5b387c96723e5f6392e069c0c5bd74fb5))
+* add vanilla-chat test agent for local evaluator testing ([#21](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/21)) ([d2ca4a2](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/d2ca4a286d5a031a2953aa3d1fe015ee82f66357))
+* added a opfor sdk as runner ([675999f](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/675999f7704822f0b68d2ecc7d0da5512742b516))
+* added mcp ([814a27a](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/814a27a71ec0d40a27e04bf35e494613d1f45e72))
+* added readme for astra ([ccc8049](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/ccc80491bd6739ca35dff1bd0dc6079613fb0d72))
+* added rug pull detection. changes to evaluator ([17c081a](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/17c081af1a062a165c5d915ed1dfebb6b52eac40))
+* added traces filter ([ba12d78](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/ba12d785d9814f4e59a087af46ff035f58fbb6ea))
+* **astra-mcp:** update html repoort generation format ([dbd71a2](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/dbd71a2d4973b80c1953d1f124bfdac2c4e504df))
+* **attacker:** add harness framing and tighten output contract ([6852b90](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/6852b909486ed7ac206220b81c0fcc28c7dbad2e))
+* **attacker:** Crescendo escalation with technique tag parsing ([#51](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/51)) ([fbb7556](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/fbb7556e10aba8853b1a7cb66bc1e3a5bafb9437))
+* auto feature in sdk ([f502bb6](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/f502bb62f2291c57088e0f10ef1ba0c434554d62))
+* **autonomous:** add business-integrity class and archetype-aware vector selection ([#64](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/64)) ([0d5d0db](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/0d5d0dbd09dca74c2efb9eb91cb96a735216cbb2))
+* **autonomous:** tree-based adaptive red-team agent ([#67](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/67)) ([1764627](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/1764627ba654546d219d554483f3762df6a40340))
+* change command name from run to execute ([e3e6944](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/e3e694416814a291944e73ad036f85107c0ebcc3))
+* change command name from run to execute ([069ed88](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/069ed88402ebeb74502643873ea68156a5cc6850))
+* chat widget finder ([f806154](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/f806154abbd127075b996690f952d2b73d1d6cf6))
+* chat widget finding enhancements ([32ccf5c](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/32ccf5ceea8bec71c4f7a13f1f425e85bc0da6a0))
+* cli fix ([6f3b112](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/6f3b11205136c8b28ee7bde1c2958d7859ec2e0e))
+* **cli:** add docs for cli ([9e83d75](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/9e83d75e595efe9c61239d7be29750d5144e80f9))
+* **cli:** add local-script target ([d8f4fec](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/d8f4fec8f28916681beb7bbabe8b5f3d7cbd5745))
+* **cli:** add progress logs to agent generate and run commands ([2723df7](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/2723df70a1a0c4ae51f46f3260643120d7b0b937))
+* **cli:** add standalone TypeScript CLI with init, setup, and run ([f142c62](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/f142c6288231a0c93c290c6b6067fadc4ba44243))
+* **cli:** opfor execute runs the setup wizard inline when --config is omitted ([e35b0fe](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/e35b0fe57ac0748440819140752ed4ccb302634c))
+* **core:** stateless multi-turn for raw LLM APIs via target.stateful flag ([#57](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/57)) ([2449622](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/244962205949eb6d200350121df16b251448a94a))
+* create pdf report generation script ([8521f29](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/8521f29ca4651b05569949d08f3debaac6630885))
+* cross session context sharing ([01bed76](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/01bed7674a9c9828b75356eedd31c6259ae53892))
+* cross session context sharing ([8501830](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/8501830bad9efe981440dae9b4b6f4c6ff6ff72a))
+* enhance adversarial defense mechanisms in prompts and tools ([4ee277a](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/4ee277a04002a08916d1694fe29b44044ffd5852))
+* enhance chat interaction capabilities and configuration management ([7a3b46d](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/7a3b46d5418e3a02f94401457face512c4fb09da))
+* Enhance error handling and reporting in run command ([7de471b](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/7de471bec03affb519633750d5be11e99dbabea5))
+* enhance evaluator selection and settings persistence ([6f998d2](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/6f998d220270881260aca9ed31574a556b9b39ec))
+* ERROR verdict for target failures, improved HTML reports, fix 422 session-id bug, cleaner CLI output ([4561049](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/45610495eb6d43d88a65828dcb7682d77ba52d28))
+* evaluator/suites refactor ([fbf09cf](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/fbf09cf979134654e5ca017beaecbda6f8bb3f0f))
+* **evaluators:** add harmful-content suite (12 harm categories) ([#60](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/60)) ([3c13154](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/3c13154e17cad4a042c4e891cecce5e5d95ecf0c))
+* extension yaml update ([fc28d53](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/fc28d534efe956cbdc883047cfef8eedb194e5a3))
+* **extension:** add All_Evaluators suite for cross-suite selection ([2751a5a](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/2751a5a6637699b465b85a9f3285e22c6b61b9c2))
+* fine tune skills ([78a214f](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/78a214fb0ff8188883df1fd3492474819c14c88a))
+* fine tune skills ([1bbd11b](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/1bbd11b159e156495be93f8ef1415bf17a87acf0))
+* fix multi turn flow, enhance report and handle errors better in run phase ([e87b77a](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/e87b77ad333f44f31e3cbfb7c7941d0231d6d4c0))
+* generate script update ([250aee4](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/250aee4ed86e0427e62615b74269c47a460163dd))
+* implement autonomous-agent ([97d2719](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/97d27193134c3b82a209da9d08284a5f947d50af))
+* implement chat widget detection and user interaction handling ([#42](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/42)) ([c036b5e](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/c036b5ebd0c9ad6535a66bb1ad49180ba0f0ae04))
+* improve agent-redteaming judge with turn citations and multi-turn modifier ([#59](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/59)) ([dcf8395](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/dcf83959ce442deccabd13a91cdf570bc3bb3dd5))
+* improve model provider UX in extension popup ([3ba4c6f](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/3ba4c6f2ff49817b5b1f5655e46eed1b0e622497))
+* improved chat reader and updated prompt injection, excessive agency prompts ([bfed377](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/bfed377d731e4e3eaf3348dea642b3f0536cc96d))
+* improved chat widget finder ([9d888a5](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/9d888a55d389198b9e9080f3ca8d41761cd4da64))
+* issue templates added ([43cc776](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/43cc776ffe597b36f5ec15684a0a71b98510857e))
+* Langfuse trace propagation, per-attack trace ids, and trace-grounded attack prompts ([5e6a806](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/5e6a806a4697f6ec0bc8d3bb799c34786dbea927))
+* **mcp:** add inline tool parameters and astra_list_evaluators tool ([d9d0510](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/d9d051031ca4c2dd3ffc2a2dd86fe7c9ccb3724c))
+* **mcp:** add multi turn for mcp ([8443ee5](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/8443ee5be1c6e4ece60ece880b94c6d5a2de8b8d))
+* **mcp:** add multi turn for mcp ([f2850d9](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/f2850d916818f419c577b2a49fca1b1e10a58f01))
+* migrate evaluator frontmatter from ref to standards and update engine/catalog/validation to use standards end-to-end ([#53](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/53)) ([13fb2d8](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/13fb2d8492355ae138c9b703f27a76b28e6d1bba))
+* multi-turn attack support, custom JSON path mapping, and session ID propagation ([f8ed8cc](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/f8ed8cc5d8370c685146407b0cf8a5111de1c270))
+* multi-turn attack support, custom JSON path mapping, and session ID propagation ([fdec26f](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/fdec26fac5779450142360cadae82817591262f2))
+* rebuild jailbreaking/goal-hijack/tool-misuse with multi-turn attacks ([cb0f0ff](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/cb0f0ff8902ab03b346140cc883cf6b85794a35b))
+* remove generated timestamp ([#80](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/80)) ([15402bc](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/15402bc8e2833e335cd939d4cc46765240501134))
+* removed enviroment: llm tests gate from workflow ([3f6cf2c](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/3f6cf2c0de06c9506f71c3ab8722e0a1bfa7e924))
+* replace holistic judge prompt with rubric-based scoring checklist and linting ([#24](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/24)) ([817e08c](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/817e08c9770533f8f4531a7c60675736fef06ea2))
+* restructure attack generation prompts with research-driven best practices ([#25](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/25)) ([e347f1f](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/e347f1fb72f6d70ff718be79344a2f5542daed85))
+* restructure attack generation prompts with research-driven best practices ([#26](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/26)) ([9123202](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/9123202ba9c80f324a59fa99dfe14cb36153193f))
+* setup for opfor mcp ([84361d8](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/84361d85ec6edaf48346a05fd704a608526ac62a))
+* setup for opfor mcp ([c907ec6](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/c907ec6159c89899d48b41178e610227f30276ba))
+* standardize LLM provider interfaces across CLI, MCP, and extension ([#37](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/37)) ([130e3da](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/130e3da50cbb43d290a6251cef038f9b1b9e38cd))
+* support different llm configs for attack and judge ([69db348](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/69db348b1513778c4abce94142556f50ca31bca0))
+* support different llm configs for attack and judge ([80af47a](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/80af47a941de52d9f55ffe36cabf9613a5f7ce02))
+* support html report generation ([34f1a40](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/34f1a409a1fc461fe58bc1d32f2170692aa08434))
+* **telemetry:** poll traces to completeness for judge; thread trace context into adaptive attacks; discover flat-file evaluators ([4e399e5](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/4e399e5659653c8fd3cb71d7efc401ea95f2f1c7))
+* tuned mcp evals and prompt gen ([7f4b6bc](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/7f4b6bce43e96b62f3149d88cba31634045e7ba6))
+* tuned mcp evals and prompt gen ([80dc62a](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/80dc62a1e528bb23858cdb9673a87b360e14c85b))
+* tuned mcp evals and prompt gen ([0b31b21](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/0b31b219c059b68a9c3213454fe780e920590ab0))
+* update astra-setup and astra-run skills with Netra telemetry, trace-grounded attack generation, attacker LLM multi-turn, and trace-enriched judge ([b4b4448](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/b4b4448fa19b48138f89cb40c227d978abbd71b8))
+* update config skill for auto-detection ([af8f7aa](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/af8f7aa2db5bff17548e2739a2315be36426bc7d))
+* update content-injection and ssrf evaluators with enhanced criteria and templates ([38b25d6](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/38b25d6a020a0d77adeeb409b6f4abfbd9f7a387))
+* update diagram with sdk ([1d48076](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/1d480766bf472dc244e9623a5f326a65593a5139))
+* update extension ui ([bbea804](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/bbea80431769a6db080b2c61e39e4f7c5af651c5))
+* update mcp red team to agent red team struct. Add skill support for mcp red team ([c989b83](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/c989b837ab6663933aa29cc766cd343ac3223ed9))
+* Update README.md ([48e7592](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/48e75920e3a0125917435e180dcff350ad5e6184))
+* update template for cli and mcp ([b7cbf23](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/b7cbf239d26035f36a6760286644122b860c46f6))
+* update testcase generation architecture ([45753bb](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/45753bb9481dd9e2b37d7a2f94a1f04d974aa9c7))
+* updated extension manifest to agent opfor ([cae3178](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/cae3178f905459ed0d1a8fd728493ad5e94dde68))
+* upgrade extension pop UI ([9fdd42b](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/9fdd42b3dcb6267fb97860c20288d805e2f27b4a))
+* workflow updated to only trigger on evaluator file change ([a694d0a](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/a694d0a19dccbea7d300e35058a9a4721b331cba))
+
+
+### Bug Fixes
+
+* add fix for red team run to detect directory ([67f3e9c](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/67f3e9c722c5a9c64ca87bfa171847895ace3744))
+* add swallowed comments to empty catches, enforce no-empty as error in extension ([e2863a1](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/e2863a158c499df25ef43b36f1e40a50acc1f459))
+* all evaluator not working bug on extension ([c585994](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/c5859947a0b1d04ed55c448848b52f39bb30283e))
+* assign finding ranks after score sort, clarify turns schema comment ([6daa049](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/6daa049532def74daa28f09b888a761ecaa5b9c4))
+* catalog file gen fix ([#85](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/85)) ([c6a52d1](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/c6a52d18fc27477a3c226850741d73e9626fc3f7))
+* Change font, minor style fixes ([ff58b9a](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/ff58b9a757728dd6b075576ce9e0d8498eba5572))
+* Change font, minor style fixes ([b0087c2](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/b0087c2cb530ed80c1ab581a6c23c833d97841c6))
+* change hardcorded trace provider names in skills to generalized terms ([5c6c587](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/5c6c58732b4083159d5b3e7251d460159ef49056))
+* change temperature for gpt 5 ([251037a](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/251037abb405b088d13148adbb9d320e985a65d0))
+* complex zod schema caused build failure for mcp ([26f8a98](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/26f8a985279382ae338c59aca958a8f4c4e88bf4))
+* **deps:** bump langchain ecosystem to v1.x in e2e test agents ([c68cf45](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/c68cf457792092d29776958574cba18be67203b5))
+* **extension:** correct phase broadcast order — show running during attacks, judging after ([4f71c69](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/4f71c6986dabda5a1bfd229d00c71c735c076810))
+* **extension:** keep service worker alive during multi-evaluator runs ([8a9fa89](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/8a9fa89f42c49a34bdbc690a43d32798f228fa77))
+* **extension:** pass LlmConfig (apiKeyEnv) not raw profile (apiKey) to runAllBrowser ([2d675aa](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/2d675aa9d87009d7237bce92c8e86bf0f48a7edb))
+* **extension:** patch DomTarget interface conformance and undefined siteUrl ([8337096](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/8337096d97bbf23baefee2edb4a62cf1b6387cb0))
+* **extension:** restore UI animation, response recovery, and temperature compat ([6b0f726](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/6b0f726fb43fc50f39dd156587d5677800384e83))
+* feed judge verdict/reasoning into multi-turn attacker history, remove stale chain-judging instruction ([7849d85](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/7849d85546150f58dd16296d98c54c3bf63fa218))
+* handle graceful exit for astra setup ([d15dff3](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/d15dff3e52275d13f1271780a4f3add1c6403e57))
+* hard cap agent response to 120k tokens ([c7cf3ed](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/c7cf3ed380a00dce90f8f7267f9dec64c1d068be))
+* **hunt:** enforce budget mid-stream, force synthesis on interrupt, default commander to sonnet ([8abe2a5](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/8abe2a52c5fa669fd8c9b2a55a2943bbae93820d))
+* **hunt:** real-time budget enforcement, forced synthesis on interrupt, default commander to sonnet ([45bc118](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/45bc1183aac0b9f6d2dac65666f33012327be787))
+* hydrate Netra trace metadata, rename SetupResult curation fields provider-agnostic ([de3b27f](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/de3b27f21f5a8bfcf71d28a774ee299080ed5598))
+* inject attack context and add evaluator scope check ([#30](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/30)) ([4ab35ee](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/4ab35ee738833a87625029ea0468ca1ed3c2739f))
+* **judge:** parse verdicts tolerantly, ERROR (not FAIL) on parse failure ([#84](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/84)) ([c90bb32](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/c90bb32787bb1b1f85dc22edbf3f14de6808690a))
+* lint issue in runner resolved ([62af852](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/62af852a7ac742840c6f85d9b9eb44a35af3d42f))
+* Making Exectue button fixed, and changed disable condtion ([a8f6cec](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/a8f6cec3116b626be4913d98da4e9b29ef775c11))
+* MCP schema field mismatch, dead code removal, pre-release hygiene ([2ebe1f0](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/2ebe1f0a620479765b020d208081e490fb82fde2))
+* MCP schema field mismatch, dead code removal, pre-release hygiene ([63286d6](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/63286d6da62396b09c7fd5a8f45f2390be14c823))
+* **mcp-astra:** fix issue with prompt gen hallucinating evaluator suite name ([27681a8](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/27681a862e4af1841fab3d0cb3d35f0892e32abd))
+* **mcp:** change mcp prompts for better UX ([f1d93a7](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/f1d93a7c7a2ff69d051e304275473dc2e6806ecf))
+* **mcp:** change mcp prompts for better UX ([2af5e9c](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/2af5e9c329fd0471ad6c14a2d1133d002645a46c))
+* **mcp:** remove stale [@ts-expect-error](https://github.com/ts-expect-error) on astra_setup tool ([f3e78f3](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/f3e78f3fb685f543d05f540fa257894bbca05cf3))
+* pass full target response to judge LLM without truncation ([986e34d](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/986e34dbe7d20450c9a002b290910a137b05c915))
+* pass per-attack sessionId to target.send for multi-turn ([77a4cb0](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/77a4cb060ea92423451a9b8c58f6cf55b304a64c))
+* process bug fix ([ac56c61](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/ac56c619606cb70a393ce7abc3ffd04db104e0a0))
+* remove depth info ([21f59fc](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/21f59fc6bd927d308fac9f6a3d5542a770fd9971))
+* rename attackerLlm, fix env override, trace path, and Langfuse baseUrl ([c749deb](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/c749deb4df5f7ebb1c509e481298504dc2383afd))
+* rename attackLlm → attackerLlm, fix env/trace/credential bugs ([389d7b3](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/389d7b3f36709d96050d86410cdbdcd24c77a516))
+* report generation failure while running tests with stdio ([d707f75](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/d707f7591c60f4a9e0e8f03167d99db404ed260b))
+* Report UI changes ([c947766](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/c9477664ec582fc21d905156e75d21c0654a62df))
+* report ui update ([6eaa216](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/6eaa216691f73a1d52c369490d83e317a2bc672e))
+* restore turnMode + telemetry wizard, per-run report folders, and CLI polish ([949e3f7](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/949e3f71732dc84fbb5faa57ce3ff0a8e0d800a6))
+* root html override fix ([9db38cd](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/9db38cd0704ca9074e1333903e4a9865b9d7b0ae))
+* run judge once after all turns, not per-turn (restore c355551 inside runAll.ts) ([2e7fc02](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/2e7fc020f0e7fddb3fe9caeddb3baa497e2c5248))
+* standardize hunt report filenames and rename run report prefix to run-report- ([5c0e13d](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/5c0e13d72480f4f7f591f05699e87cac8cb8471f))
+* UI / UX fixes ([#54](https://github.com/KeyValueSoftwareSystems/agent-opfor/issues/54)) ([79babc2](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/79babc25dee5199564dab23c624560d08799ddb1))
+* Ui updations in extention, Style changes ([3e5166e](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/3e5166e2fb46108155a2e6b02eb2b0df2299624b))
+* unify LlmConfig, validate run config, discriminate attack types, dedupe report math ([b6e384a](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/b6e384a86ea08262754461c4bf1c2e32f6ac2a7c))
+* update generatedAt timestamp and improve dropdown styling ([817b507](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/817b5071e6e1f7ecb87eb78a97c5f5baf1f938d5))
+* update red team config to avoid hallucination for choose of suites ([7ded8ce](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/7ded8ce6951abf4a4a74e2317d634972bbc73008))
+* update report generation code for better readability in turnwise dropdown ([ada5049](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/ada504914bebd262a21b0b15950d13a97a20fb0a))
+* widen test tsconfig rootDir to resolve cross directory imports ([9ba9d09](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/9ba9d09f7e2c03df01271ddfde2cd56bccf80e93))
+* wire MCP evaluator catalog and expand env vars in URL transport headers ([70b7a70](https://github.com/KeyValueSoftwareSystems/agent-opfor/commit/70b7a70a3bb51401cc9e10c10ef32fbce3006f0b))
diff --git a/core/package.json b/core/package.json
index 2bf08342..e602dc9f 100644
--- a/core/package.json
+++ b/core/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@agent-opfor/core",
-  "version": "0.9.0",
+  "version": "0.9.1",
   "description": "Opfor core engine — attacker prompt generation, judge, and execution shared by all runners",
   "license": "Apache-2.0",
   "private": true,
diff --git a/package-lock.json b/package-lock.json
index 41588dee..57290452 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -1,12 +1,12 @@
 {
   "name": "agent-opfor",
-  "version": "0.1.0",
+  "version": "0.9.1",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "agent-opfor",
-      "version": "0.1.0",
+      "version": "0.9.1",
       "license": "Apache-2.0",
       "workspaces": [
         "core",
diff --git a/package.json b/package.json
index 8208543f..de639b44 100644
--- a/package.json
+++ b/package.json
@@ -1,6 +1,6 @@
 {
   "name": "agent-opfor",
-  "version": "0.9.0",
+  "version": "0.9.1",
   "description": "Opfor — security testing for AI agents and MCP servers (workspace root)",
   "license": "Apache-2.0",
   "private": true,
diff --git a/runners/cli/package.json b/runners/cli/package.json
index 8c0f517e..ec21ae4a 100644
--- a/runners/cli/package.json
+++ b/runners/cli/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@agent-opfor/cli",
-  "version": "0.9.0",
+  "version": "0.9.1",
   "description": "Opfor CLI — security testing for AI agents and MCP servers (opfor setup|run|hunt)",
   "license": "Apache-2.0",
   "type": "module",
diff --git a/runners/cli/ui/package.json b/runners/cli/ui/package.json
index 8c7963a7..4c1b0505 100644
--- a/runners/cli/ui/package.json
+++ b/runners/cli/ui/package.json
@@ -1,7 +1,7 @@
 {
   "name": "@agent-opfor/autonomous-ui",
   "private": true,
-  "version": "0.9.0",
+  "version": "0.9.1",
   "type": "module",
   "scripts": {
     "dev": "vite",
diff --git a/runners/extension/package.json b/runners/extension/package.json
index cae6c0b9..bfa639a3 100644
--- a/runners/extension/package.json
+++ b/runners/extension/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@agent-opfor/extension",
-  "version": "0.9.0",
+  "version": "0.9.1",
   "description": "Opfor browser extension (MV3) — chat UI injector for live testing",
   "license": "Apache-2.0",
   "private": true,
diff --git a/runners/mcp/package.json b/runners/mcp/package.json
index e0b0ca2e..e635c4ed 100644
--- a/runners/mcp/package.json
+++ b/runners/mcp/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@agent-opfor/mcp",
-  "version": "0.9.0",
+  "version": "0.9.1",
   "description": "Opfor MCP server — expose red team tools to any MCP-compatible AI agent",
   "license": "Apache-2.0",
   "type": "module",
diff --git a/runners/sdk/package.json b/runners/sdk/package.json
index f3fe1526..cf258483 100644
--- a/runners/sdk/package.json
+++ b/runners/sdk/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@agent-opfor/sdk",
-  "version": "0.9.0",
+  "version": "0.9.1",
   "description": "Opfor SDK — programmatic adversarial testing for AI systems",
   "license": "Apache-2.0",
   "type": "module",
diff --git a/tests/e2e/agents/customer-support/package.json b/tests/e2e/agents/customer-support/package.json
index 95f99d25..a291886d 100644
--- a/tests/e2e/agents/customer-support/package.json
+++ b/tests/e2e/agents/customer-support/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@agent-opfor/test-agent-customer-support",
-  "version": "0.9.0",
+  "version": "0.9.1",
   "description": "Customer support test agent with PostgreSQL — tests BOLA, BFLA, RBAC, PII, and SQL injection evaluators",
   "private": true,
   "type": "module",
diff --git a/tests/e2e/agents/vanilla-chat/package.json b/tests/e2e/agents/vanilla-chat/package.json
index 1339e4f0..851f6b25 100644
--- a/tests/e2e/agents/vanilla-chat/package.json
+++ b/tests/e2e/agents/vanilla-chat/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@agent-opfor/test-agent-vanilla-chat",
-  "version": "0.9.0",
+  "version": "0.9.1",
   "description": "Vanilla chat test agent — used for local developer testing of Opfor evaluators",
   "private": true,
   "type": "module",
diff --git a/tests/e2e/agents/vulnerable-memory/package.json b/tests/e2e/agents/vulnerable-memory/package.json
index 21265537..536b0ef9 100644
--- a/tests/e2e/agents/vulnerable-memory/package.json
+++ b/tests/e2e/agents/vulnerable-memory/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@agent-opfor/test-agent-vulnerable-memory",
-  "version": "0.9.0",
+  "version": "0.9.1",
   "description": "Intentionally vulnerable agent — accepts and persists user-injected 'policies' into a global knowledge base across sessions",
   "private": true,
   "type": "module",
diff --git a/tests/e2e/mcp/vulnerable-server/package.json b/tests/e2e/mcp/vulnerable-server/package.json
index 99e860e5..ce5cec22 100644
--- a/tests/e2e/mcp/vulnerable-server/package.json
+++ b/tests/e2e/mcp/vulnerable-server/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@agent-opfor/test-mcp-vulnerable",
-  "version": "0.9.0",
+  "version": "0.9.1",
   "description": "Intentionally vulnerable MCP server for opfor testing — DO NOT deploy",
   "private": true,
   "type": "module",