Commit Graph

25 Commits

Author SHA1 Message Date
ollie 42361e6505 execute: support elevated+detach combination
Rewrite executeElevated to use the same select/detachCh pattern as
executeWithStdin. When both elevated and detach are set, the process
is immediately backgrounded into the detach registry with streaming
output captured in the ring buffer.

Also adds the detach field to the execute_code JSON schema description.
2026-07-18 00:16:20 +02:00
ollie c1124b588d execute: add detach field for autonomous background processes
When detach: true is set on a CodeStep, the process is immediately
backgrounded after start. Its stdout/stderr are captured in a 64KB
ring buffer accessible via GetDetachedOutput. The agent can then
inspect and manage the process via process_list/process_output tools.
2026-07-17 23:28:14 +02:00
Levi Neely 16d057741c execute: add process detach mechanism
Allows the user to detach a running process from the agent mid-execution.
The process continues running; output flows to a 64KB ring buffer.
Detached processes can be listed, signaled (TERM/KILL), and dismissed.

- DetachedProcess struct with ring buffer, signal, output methods
- Server.Detach() signals the currently executing process to detach
- Execute() rewritten: cmd.Start() + select on wait/ctx/detach
- cleanupDetached() on Close() sends SIGTERM to all
- OnDetach/OnExit hooks for integration with D-Bus layer
2026-07-17 16:43:07 +02:00
ollie a50149d075 fix: send profileArn for social accounts, improve error messages
- Remove IsSocial guard from sqliteKiroAuth.ProfileARN(); the Kiro API
  now requires profileArn on all requests including personal accounts
- Add OLLIE_KIRO_PROFILE_ARN env var for static token auth
- Add repeated error detection (3x identical error triggers corrective nudge)
- Include expected format in tool dispatch error messages
2026-07-11 16:08:44 +02:00
Levi Neely fd20070294 refactor: remove 9P mount logic from core execute server
Moved all 9P FUSE mount code from core to 9p module using lifecycle hooks.

Changes:
- Removed 9P-specific fields (mountProc, mountPath, mountEnabled) from execute.Server
- Removed WithMount(), MountPath(), startMount(), remountIfStale() functions
- Added lifecycle hooks: OnPreDispatch, OnClose, OnEnvSet with corresponding options
- Updated SetEnv(), Close(), and Dispatch() to call hooks instead of hardcoded mount logic
- Fixed tool descriptions to use $OLLIE_TOOLS_PATH instead of 9P-specific $OLLIE/s/$SID/t paths
- Removed unused execute import from agent/commands.go

The core execute server is now harness-agnostic.
2026-07-08 15:55:10 +02:00
Levi Neely 4188d5946b execute: transparent per-session mount recovery on stale FUSE 2026-05-21 08:49:36 +02:00
Levi Neely 1207e26a07 remove trustedTools and Confirm callback; trust is implicit via allowExecutors 2026-05-17 15:21:26 +02:00
Levi Neely 161d0d237f add allowExecutors/allowTools two-layer tool restriction 2026-05-16 23:54:55 +02:00
Levi Neely 6dfae368d0 fix: wire PlanFile, add missing rate-limit constants, simplify test helpers, fix regex
- Wire PlanFile from $OLLIE/s/{id}/plan into agentConfig so plan
  re-injection is no longer dead code
- Add missing failureWindow/maxFailures/blockDuration constants
- Replace hand-rolled contains() with strings.Contains
- Fix rm -rf .. regex to match end-of-string and whitespace
- Add TestPlanReinjection covering periodic plan re-surfacing
- Add loop_maxsteps_helpers_test.go with mockResponse/newState/sequentialStream
2026-05-16 23:54:04 +02:00
Levi Neely 0903254c6c execute: wire call_tool and pipe dispatch; execute_code is code-only 2026-05-15 18:48:23 +00:00
Levi Neely c02607aa60 execute: separate steps (parallel) from pipe (sequential)
- steps: independent, run in parallel when annotated safe, outputs
  concatenated. No stdout chaining.
- pipe: explicit pipeline, sequential, stdout of each stage feeds
  stdin of the next.

steps and pipe are mutually exclusive in a single execute_code call.
2026-05-11 13:50:01 +02:00
Levi Neely d638c2e75b Revert "execute: remove auto-batching from pipeline, fan-out is always explicit"
This reverts commit 2117f14a53.
2026-05-11 13:43:13 +02:00
Levi Neely 2117f14a53 execute: remove auto-batching from pipeline, fan-out is always explicit
steps is always a pipe (sequential, stdout chains to next stdin).
Concurrent fan-out requires explicit use of the parallel field.
The ollie:parallel annotation still drives agent-level tool call
parallelism via IsParallelRead, but no longer hijacks pipe semantics.
2026-05-11 13:40:47 +02:00
Logan Neely cafc55e038 execute: add Strict and Yolo server options
Strict rejects inline {code} steps, allowing only named {tool} steps.
Yolo skips the landrun sandbox entirely.
2026-05-07 22:27:22 +02:00
Levi Neely 6abf0cdb3d execute: parallel dispatch via flock; agent: preTool/postTool hooks, turn rollback, mid-turn compact
execute_code auto-batches consecutive tool steps annotated '# ollie:parallel read'
into goroutine fan-outs. Write/global steps serialize. LOCK_SH/LOCK_EX via
syscall.Flock in a per-session lock dir for cross-process safety. Tool scripts
declare their own parallelism profile; no tool names hardcoded in Go.

- flock.go: lockClass, classifyStep (annotation-driven), acquireFlock
- tools.go: LockDirSetter interface
- server.go: lockDir field, SetLockDir method
- pipe.go: runReadBatch with shared-lock goroutine fan-out
- code.go: auto-batching pipeline loop
- core.go: pushLockDir wires session tmpdir to execute server at init;
  preTool/postTool hook callbacks in agentConfig; AutoCompact after each
  tool round; session snapshot/restore on turn failure

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-26 09:00:29 +02:00
Levi Neely 46272f22a6 test: sandbox coverage + backend contract tests
- sandbox: remove dead code (DefaultConfig, Load/Save, LayeredConfig, Merge,
  Version, SystemDefaults, validateName, ConfigPath, OpenSandbox); unexport
  internal symbols; LoadSandbox now takes io.Reader; loadSandboxConfig owns
  file opening; 83% package coverage via new sandbox_test.go
- backend: add BaseURL field to AnthropicBackend for testability; new
  backend_test.go runs checkContract against all five implementations
  (Ollama, OpenAI, Copilot, Anthropic, CodeWhisperer) via httptest servers

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-21 13:00:06 +02:00
Levi Neely e09a6c1250 elevation: add Elevator interface + superpowers backend; add elevated:true to execute_code steps
Remove session-scoped env injection from frontend (s/{id}/env, -e/-env flags) — superseded
by per-invocation superpowers run-session in the elevator. Keep agentCore env map and
OLLIE_SESSION_ID propagation via pushSessionEnv.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-18 20:06:43 +02:00
Levi Neely c2ba1728e4 execute: unify execute_pipe into execute_code as pipeline primitive
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-16 21:36:22 +02:00
Levi Neely ab96de2771 execute: remove execute_tool; use execute_code {tool} steps instead
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-16 21:18:12 +02:00
Levi Neely cf25268d80 execute: refactor execute_code to parallel steps; add parallel pipe stages
- execute_code now takes steps:[{code,language}|{tool,args}] instead of
  a single code/language pair. Single-step calls return raw output;
  multi-step calls run in parallel and return results under === step N ===
  headers in submission order.
- execute_pipe stages may now include {parallel:[...]} to fan out
  concurrently within a pipeline. Parallel outputs are concatenated in
  submission order and fed as stdin to the next stage.
- Stage execution is now Go-level (stdin chaining) rather than a single
  bash pipeline string, enabling parallel stages and tool/code mixing.
- CodeStep and resolveCodeStep are the shared primitives for both
  execute_code and execute_pipe parallel stages.
- executeWithStdin added to Server for stdin chaining between stages.
- Tool steps use shebang detection; language field applies only to
  inline code steps (default: bash).
- ListTools schemas and descriptions updated throughout.
2026-04-16 20:06:05 +02:00
Levi Neely 13590a84e7 implement TrustedTools: explicit trust replaces implicit allow-all
- Add TrustedToolsSetter interface and SetTrustedTools to execute.Server
- Add allowed(): trusted tools return true; untrusted with no Confirm
  return false (deny by default). Confirm returns true=allow, consistent
  with allowed() semantics throughout.
- BuildAgentEnv wires cfg.TrustedTools into the execute server

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-14 17:52:37 +02:00
Levi Neely 5d86326edf execute: language-specific dangerous patterns; skip domain-specific tools
universalPatterns: mkfs, dd if=/dev/, sudo/su, /etc/shadow|sudoers
bashPatterns: rm -rf flag syntax, fork bomb, redirects, eval "...$"
pythonPatterns: shutil.rmtree, os.system/subprocess with rm -rf, os.remove on sys paths
perlPatterns: system/exec/backtick with rm -rf, unlink glob on sys paths
luaPatterns: os.execute/io.popen with rm -rf, os.remove on sys paths
Domain-specific tools (awk, sed, jq, ed, bc, expect) skip validation.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-14 13:36:18 +02:00
Levi Neely 2d122c70dd gate file tools with sandbox path checks
file_read, file_write, file_edit, file_grep, and file_glob now check
paths against the sandbox config before operating. Same policy that
landrun enforces on execute_code, applied at the application level.

- Add sandbox.LoadMerged: loads base config + named layer, graceful
  fallback if layer missing. Replaces duplicated logic in execute.
- Add sandbox.CheckPath: resolves symlinks, expands templates, checks
  path against allowed entries (RO/ROX for read, RW/RWX for write).
- file.Server.checkAccess wires the check into every file dispatch.
2026-04-13 18:37:17 +02:00
Levi Neely 1b44244f81 cleanup tools 2026-04-12 16:26:50 +02:00
Levi Neely a3e1a078c4 refactor: move mcp/tools/execute to pkg/ for external consumability
internal/mcp      → pkg/mcp         (MCP client, concrete)
internal/tools    → pkg/tools        (Executor interface + MCPExecutor)
internal/exec     → pkg/tools/execute (builtin sandbox executor, split by tool)

Split exec.go into executor.go (sandbox runner), pipe.go (BuildPipeline/PipeStep),
tool.go (ReadTool), and code.go (execute_code is Executor.Execute with trusted=false).

All three packages are now importable by consumers outside this module.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-09 21:59:01 +02:00