Add ChatGPT plugin cookbook with Agents SDK and AgentCore

Author: rohan-eliza

examples/partners/AWS/chatgpt_agents_sdk_aws_agentcore_cookbook/.env.example

@@ -0,0 +1,77 @@
+# Copy to .env for local development. Do not commit secrets.
+
+AWS_REGION=us-west-2
+OPENAI_BASE_URL=https://bedrock-mantle.us-west-2.api.aws/v1
+OPENAI_API_KEY=
+OPENAI_AGENTS_MODEL=openai.gpt-oss-120b
+# Optional deterministic override. If empty, demo flows use a UTC date 45 days
+# in the future so instructions and smoke payloads do not expire.
+COOKBOOK_DEMO_TRAVEL_DATE=
+
+# Canonical execution switch: local runs the checked-in Python agent; deployed
+# invokes the AgentCore Runtime ARN below. Legacy FLIGHT_DATA_SOURCE values are
+# accepted for compatibility but must not conflict with this setting.
+COOKBOOK_EXECUTION_MODE=local
+LOCAL_AGENT_SERVICE_NAME=chatgpt-agentcore-cookbook-local
+LOCAL_AGENT_LOG_GROUP=/aws/bedrock-agentcore/runtimes/chatgpt-agentcore-cookbook-local
+# Optional deployed-Runtime configuration. Leave this empty for the default
+# local path. When deployed mode is selected, set it to the Region in the
+# supplied Runtime ARN; it takes precedence over AWS_REGION for that invocation.
+AGENTCORE_RUNTIME_REGION=
+AGENTCORE_RUNTIME_AGENT_ARN=
+AGENTCORE_RUNTIME_QUALIFIER=
+# Set only when this value is derived from an authenticated principal.
+AGENTCORE_RUNTIME_USER_ID=
+
+# Optional live AgentCore evaluation. This is separate from the offline
+# Promptfoo suite. Keep the guard disabled unless you intend to invoke AWS and
+# accept the resulting usage charges. AWS_PROFILE belongs in your shell, not
+# this file.
+RUN_AWS_EVALUATION=0
+AGENTCORE_EVALUATION_LOG_GROUP=
+# The built-in evaluators need prompt, response, and tool content in the spans.
+# Set this to 1 only for a dedicated evaluation Runtime whose owner has enabled
+# and approved OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT=true.
+AGENTCORE_EVALUATION_CONTENT_CAPTURE_CONFIRMED=0
+# Optional tuning. These values show the runner defaults.
+AGENTCORE_EVALUATION_WAIT_SECONDS=180
+AGENTCORE_EVALUATION_POLL_SECONDS=30
+AGENTCORE_EVALUATION_ATTEMPTS=5
+AGENTCORE_EVALUATION_MIN_SCORE=0.5
+# Optional comma-separated case IDs. Empty means every AWS-tagged canonical case.
+AGENTCORE_EVALUATION_CASE_IDS=
+
+# Primary pre-promotion evaluation over actual local Agents SDK outputs.
+# Keep the guard disabled in .env and enable it only for the intended command.
+RUN_PROMPTFOO_AGENT_EVALUATION=0
+# Optional comma-separated runnable case IDs. Empty means all three canonical cases.
+PROMPTFOO_AGENT_EVALUATION_CASE_IDS=
+PROMPTFOO_AGENT_EVALUATION_TIMEOUT_MS=180000
+
+# Trace destinations. AWS is the default and requires no OpenAI trace key.
+# Set COOKBOOK_TRACING_MODE=dual only after a separate OpenAI Platform project
+# key has been approved for trace export.
+COOKBOOK_TRACING_MODE=aws
+# Optional routing override and credential, used only in explicit dual mode.
+OPENAI_PROJECT_ID=
+OPENAI_TRACE_API_KEY=
+# Optional base name; the runtime always appends `(local)` or `(deployed)`.
+OPENAI_TRACE_WORKFLOW_NAME="ChatGPT flight agent"
+# Keep prompt, model output, and tool payload content out of traces by default.
+OPENAI_TRACE_INCLUDE_SENSITIVE_DATA=0
+# Apply the same content policy to the ADOT/OpenTelemetry exporter.
+OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT=false
+DISABLE_ADOT_OBSERVABILITY=false
+OTEL_SDK_DISABLED=false
+# The local invoker sets this automatically to prevent auto-instrumentation
+# from constructing the OpenAI trace exporter before the AWS-only setup runs.
+OTEL_PYTHON_DISABLED_INSTRUMENTATIONS=
+
+# Bounded, read-only trace verification defaults. These do not prove a trace
+# exists until verify_traces.py reports AWS status "verified".
+COOKBOOK_TRACE_VERIFICATION_LOG_GROUP=aws/spans
+COOKBOOK_TRACE_VERIFY_TIMEOUT_SECONDS=60
+COOKBOOK_TRACE_VERIFY_POLL_SECONDS=5
+
+# Notebook gates. Local/test cells run by default; live AWS cells require RUN_LIVE_AWS=1.
+RUN_LIVE_AWS=0

examples/partners/AWS/chatgpt_agents_sdk_aws_agentcore_cookbook/.gitignore

@@ -0,0 +1,13 @@
+.DS_Store
+.env
+.venv/
+__pycache__/
+.pytest_cache/
+.ruff_cache/
+node_modules/
+dist/
+runtime-agent/evals/results/
+*.log
+*.tmp
+notebooks/*.executed.ipynb
+notebooks/.ipynb_checkpoints/

examples/partners/AWS/chatgpt_agents_sdk_aws_agentcore_cookbook/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Eliza
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.