adding a write result tool - #399
Closed
DhavalRepo18 wants to merge 796 commits into
Closed
Conversation
The LLM was guessing argument names (e.g. 'site' instead of 'site_name', 'asset_name' instead of 'asset_id') because the prompt gave no schema. execute_plan now pre-fetches tool schemas for all referenced servers (one _list_tools call per server, reusing the MCP connection), then passes each tool's parameter signature to execute_step -> _resolve_args_with_llm. The prompt instructs the LLM to use EXACTLY the listed parameter names. New tests: schema appears in prompt, empty schema renders as '(unknown)'. Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
_parse_json returned {} for both "parsed empty object" and "parse failure",
so the warning in _resolve_args_with_llm fired for tools like
current_date_time that legitimately take no arguments.
_parse_json now returns None on failure; {} is a valid successful result.
The warning only triggers on None, and the caller returns {} as fallback.
Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
assets/sensors/history now include the actual values in their message field so downstream steps (and the LLM) have the data inline without having to parse the full response object. Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
The planner's server description format is "tool(param: type): desc".
The LLM sometimes echoes the full signature in #Tool lines, e.g.
"#Tool2: assets(site_name: string)", which caused MCP to warn "tool not
listed" and skip validation. parse_plan now splits on "(" and keeps
only the bare tool name.
Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
- Move src/workflow/ → src/agent/ - Update all internal imports (from workflow.* → from agent.*) - Update patch targets in tests (workflow.executor.* → agent.executor.*) - Update pyproject.toml: packages list and plan-execute entry point - Update INSTRUCTIONS.md and CLAUDE.md references Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
AgentRunner defines the common interface (run method + llm attribute) that all future agent implementations must satisfy. PlanExecuteRunner now inherits from it and calls super().__init__(llm). Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
Move executor, planner, runner, and models (Plan, PlanStep, StepResult) into src/agent/plan_execute/. AgentRunner ABC and OrchestratorResult remain in src/agent/ as the shared orchestration contract. Update all imports and patch targets in tests accordingly. Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
fix: strip tool signature suffix copied by LLM from server descriptions
Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
Update README.md
Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
…ays" This reverts commit 8e125b8. Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
…low-to-agent Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
… mode list Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
chore: add stale bot to auto-close inactive issues
refactor: rename workflow to agent
Docs guidelines
Closes #244 - Add `src/agent/claude_agent/` subpackage with `ClaudeAgentRunner` that implements `AgentRunner` via the claude-agent-sdk agentic loop - Wire the same IoT/FMSR/TSFM/utilities/WO MCP servers as stdio servers via `ClaudeAgentOptions.mcp_servers` - Add `claude-agent` CLI entry point (`agent.claude_agent.cli:main`) - Add `claude-agent-sdk>=0.0.14` to project dependencies - Export `ClaudeAgentRunner` from `agent.__init__` - Add 7 unit tests (all pass, no real API calls) Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
…_execute/models - Delete src/agent/models.py - Add AgentResult to src/agent/plan_execute/models.py alongside Plan/PlanStep/StepResult - Update AgentRunner ABC and all subclasses to use AgentResult - Export AgentResult from agent.__init__ Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
…stratorResult for plan-execute - src/agent/models.py: AgentResult(question, answer, history: Any) — thin base result for all AgentRunner subclasses; history type TBD - src/agent/plan_execute/models.py: OrchestratorResult(question, answer, plan, history: list[StepResult]) — kept for PlanExecuteRunner with full plan/step-result detail - ClaudeAgentRunner.run() returns AgentResult(history=None) - AgentRunner ABC return type is AgentResult Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
…OPIC_BASE_URL Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
…tellm_proxy/ model ID Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
Signed-off-by: TokenRouter <support@tokenrouter.com>
…cture-support docs: add TokenRouter infrastructure support acknowledgment
Adding a Secutiry Check
Expose failure_code through both the update_workorder core function and the MCP tool wrapper, and cover partial-update behavior in the lifecycle test. Signed-off-by: Shuxin Lin <linshuhsin@gmail.com>
Signed-off-by: Dhaval Patel <pateldha@us.ibm.com>
adding link to arxiv paper
feat(wo): support failure_code in update_workorder
Signed-off-by: ulises-jeremias <ulisescf.24@gmail.com>
Signed-off-by: Parman Mohammadalizadeh <prmma23@gmail.com>
chore: apply ruff formatting
Update actions/checkout action to v6
…uide docs(contrib): add external industrial dataset mappings
Merger Pull Request #375
Signed-off-by: Chathurangi Shyalika <chathurangishyalika@Chathurangis-MacBook-Pro.local>
Signed-off-by: Chathurangi Shyalika <chathurangishyalika@Chathurangis-MacBook-Pro.local>
Update actions/checkout action to v7
Signed-off-by: Chathurangi Shyalika <chathurangishyalika@Chathurangis-MacBook-Pro.local>
Signed-off-by: Chathurangi Shyalika <chathurangishyalika@Chathurangis-MacBook-Pro.local>
feat(benchmark): add scenario suite runner
Signed-off-by: Chathurangi Shyalika <chathurangishyalika@daniels-mbp.watson.ibm.com>
Updating collections.json
Signed-off-by: Dhaval Patel <pateldha@us.ibm.com>
Signed-off-by: Dhaval Patel <pateldha@us.ibm.com>
Collaborator
Author
|
As discussed, I am enabling this write result tool. We will learn how to mention this in the System Prompt, and it should enforce it. |
Signed-off-by: Dhaval Patel <pateldha@us.ibm.com>
DhavalRepo18
force-pushed
the
write_db_result
branch
from
September 9, 2026 03:54
32e1c60 to
ced5795
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
19 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
issue #395