Functions — github.com/FareedKhan-dev/long-running-agent

hub / github.com/FareedKhan-dev/long-running-agent / functions

Functions438 in github.com/FareedKhan-dev/long-running-agent

⨍ Functions 438◇ Types & classes 148

↓ 2 callersMethodrequest

(self, req: GateRequest)

src/lra/hitl/gate.py:43

↓ 2 callersFunctionresearch_fanout

Investigate ``queries`` in parallel; return one brief per query (failures filtered out).

src/lra/agents/team.py:19

↓ 2 callersMethodresolve

Replace any known placeholders in ``value`` with their real secrets.

src/lra/safety/egress.py:35

↓ 2 callersMethodresolve_headers

Resolve placeholders across all header values (e.g. an Authorization header).

src/lra/safety/egress.py:41

↓ 2 callersMethodrun_mission

( self, *, workdir: str, title: str, description: str, checkli

src/lra/agents/orchestrator.py:52

↓ 2 callersFunctionrun_mission_local

Initialize the anchor and run cycles until done / blocked / over-budget / looping.

src/lra/agent/runner.py:46

↓ 2 callersFunctionrun_worker

Connect to the configured Temporal server and serve missions until cancelled.

src/lra/durable/worker.py:59

↓ 2 callersFunctionshould_declare_impossible

True when an item has failed verification too many times in a row (escalate to a human).

src/lra/ops/lifecycle.py:20

↓ 2 callersMethodspecs

The specs of tools currently permitted (what the model is told it can call).

src/lra/contracts/tools.py:72

↓ 2 callersMethodto_jsonl

Serialize the collected trace as newline-delimited JSON (for inspection/export).

src/lra/obs/events.py:61

↓ 2 callersFunctionto_openai_tools

OpenAI / OpenAI-compatible `tools` array (function-calling).

src/lra/model/tool_schemas.py:15

↓ 2 callersMethodupsert

(self, ticket: Ticket)

src/lra/durable/ledgers.py:26

↓ 2 callersMethodverify

(self, workdir: str, checks: list[Check])

src/lra/verify/verifier.py:21

↓ 2 callersMethodwrite_file

Write ``content`` to ``relpath`` (relative to ``workdir``), creating parents.

src/lra/contracts/sandbox.py:54

↓ 1 callersFunction_checklist

(n: int)

tests/durability/test_durable_spine.py:30

↓ 1 callersMethod_checkpoint

( self, cycle_id: str, item_id: str, description: str, verification: V

src/lra/agent/loop.py:189

↓ 1 callersMethod_embed_one

(self, text: str)

src/lra/memory/embeddings.py:31

↓ 1 callersFunction_extract_json

(text: str)

src/lra/agent/loop.py:50

↓ 1 callersFunction_extract_obj

(text: str)

src/lra/agents/judge.py:24

↓ 1 callersFunction_html_to_text

(html: str)

src/lra/execution/tools/web.py:123

↓ 1 callersFunction_missing_required

Return required parameter names (per JSON Schema) absent from ``arguments``.

src/lra/execution/dispatcher.py:61

↓ 1 callersFunction_parse_facts

(text: str)

src/lra/memory/consolidation.py:24

↓ 1 callersFunction_parse_tool_calls

Map OpenAI ``tool_calls`` blocks to our ``ToolCall`` (arguments arrive as a JSON string).

src/lra/model/openai_compat.py:86

↓ 1 callersMethod_read_checklist

(self)

src/lra/state/mission_anchor.py:122

↓ 1 callersMethod_read_recent_decisions

(self, n: int)

src/lra/state/mission_anchor.py:128

↓ 1 callersMethod_terminal

( self, inp: MissionInput, state: MissionState, *, completed: bool )

src/lra/durable/workflows.py:88

↓ 1 callersFunction_tracer

()

src/lra/obs/otel.py:15

↓ 1 callersMethodadd

(self, records: list[MemoryRecord])

src/lra/memory/hybrid.py:83

↓ 1 callersFunctionagent_span

Open a GenAI span named ``name`` with ``gen_ai.*`` attributes (no-op if OTel is absent).

src/lra/obs/otel.py:24

↓ 1 callersFunctionbuild_data_converter

Return the default DataConverter augmented with the ClaimCheck payload codec.

src/lra/durable/data_converter.py:17

↓ 1 callersFunctionbuild_messages

Build the initial messages for one cycle: anchor recitation + active item + tools.

src/lra/agent/prompt.py:31

↓ 1 callersFunctionbuild_worker

Construct a Worker that hosts the mission + sub-agent workflows and their activities.

src/lra/durable/worker.py:34

↓ 1 callersFunctionclaude_model_for

Map a role's model tier to a concrete Claude model id.

src/lra/agents/roles.py:121

↓ 1 callersMethodcommit_checkpoint

Atomically write the updated anchor and commit it; return the new commit sha.

src/lra/contracts/state.py:121

↓ 1 callersMethodcommit_round

Promote this round's responses into the main board and clear the response board.

src/lra/coordination/blackboard.py:38

↓ 1 callersMethodcomplete

( self, messages: list[ModelMessage], *, tools: list[dict[str, object]] | None

src/lra/model/claude.py:44

↓ 1 callersMethodestimate_cost_usd

(self, usage: Usage)

src/lra/model/claude.py:104

↓ 1 callersMethodestimate_cost_usd

(self, usage: Usage)

src/lra/model/openai_compat.py:79

↓ 1 callersFunctionevolve_and_gate

Propose an improved prompt and promote it only if it beats baseline by ``min_improvement``.

src/lra/agents/evolution.py:27

↓ 1 callersMethodexec

( self, argv: list[str], *, timeout_s: int = 600, cwd: str | None = No

src/lra/execution/sandbox_e2b.py:23

↓ 1 callersMethodget

(self, key: str)

src/lra/persistence/object_store.py:45

↓ 1 callersFunctionhead_sha

Return the current HEAD sha, or '' if there are no commits yet.

src/lra/state/git_ops.py:69

↓ 1 callersFunctionis_repo

True if ``cwd`` is inside a git work tree.

src/lra/state/git_ops.py:31

↓ 1 callersMethodmark_flaky

(self, name: str)

src/lra/verify/flaky_quarantine.py:16

↓ 1 callersFunctionmigrate

Apply registered migrators until ``data`` reaches ``target_version``.

src/lra/state/migrations.py:39

↓ 1 callersMethodnext_actionable

Return the first open item whose dependencies are all done, else None.

src/lra/contracts/state.py:44

↓ 1 callersMethodopen

(self, *, workdir: str, snapshot_id: str | None = None)

src/lra/execution/sandbox_docker.py:89

↓ 1 callersMethodopen_tickets

(self)

src/lra/durable/ledgers.py:29

↓ 1 callersMethodowner_of

(self, path: str)

src/lra/coordination/ownership.py:75

↓ 1 callersMethodpartition

Split into (gating, quarantined); quarantined checks are forced non-gating.

src/lra/verify/flaky_quarantine.py:22

↓ 1 callersFunctionplan_and_run_local

Mission intake → execution: the Planner decomposes ``task`` into a checklist, then run it.

src/lra/agent/runner.py:131

↓ 1 callersMethodpropose

( self, *, role: str, current_prompt: str, failure_traces: list[str] )

src/lra/agents/evolver.py:29

↓ 1 callersMethodput

Store ``data`` and return its content-addressed key.

src/lra/persistence/object_store.py:18

↓ 1 callersMethodquery

(self, text: str, *, k: int = 5, candidate_k: int = 20)

src/lra/memory/hybrid.py:89

↓ 1 callersMethodread

(self)

src/lra/coordination/decision_log.py:25

↓ 1 callersMethodread_file

Read ``relpath`` (relative to ``workdir``).

src/lra/contracts/sandbox.py:58

↓ 1 callersMethodread_file

(self, relpath: str)

src/lra/execution/sandbox_e2b.py:44

↓ 1 callersFunctionreconcile_in_flight

For each (ticket_id, branch) still in progress, decide adopt vs re-spawn from real git.

src/lra/durable/reconcile.py:24

↓ 1 callersMethodregister

Bind a placeholder (which the agent may see) to a real secret (which it never sees).

src/lra/safety/egress.py:31

↓ 1 callersMethodrerank

( self, query: str, hits: list[RetrievalHit], *, k: int | None = None )

src/lra/memory/rerank.py:21

↓ 1 callersMethodreset

(self, signature: str | None = None)

src/lra/governor/governor.py:71

↓ 1 callersMethodrespond

Write to the per-round response board (not yet visible on the main board).

src/lra/coordination/blackboard.py:28

↓ 1 callersMethodreview

(self, *, diff: str, criteria: str, ctx: ToolContext)

src/lra/agents/reviewer.py:33

↓ 1 callersMethodrun

(self, inp: MissionInput, state: MissionState | None = None)

src/lra/durable/workflows.py:49

↓ 1 callersMethodsnapshot

(self, session: SandboxSession)

src/lra/execution/sandbox_local.py:64

↓ 1 callersFunctionsoft_invalidate

Soft-forget: mark matching records invalid (never deletes; never rewrites). Returns count.

src/lra/memory/consolidation.py:61

↓ 1 callersMethodspecs

(self)

src/lra/execution/dispatcher.py:37

↓ 1 callersMethodtail

(self, n: int = 50)

src/lra/memory/episodic.py:19

↓ 1 callersFunctionto_claude_tools

Anthropic Messages API `tools` array.

src/lra/model/tool_schemas.py:30

↓ 1 callersMethodtransition

Return a copy advanced to ``to`` (transitions are durable events upstream).

src/lra/coordination/ticket.py:49

↓ 1 callersMethodverify

Run ``checks`` in ``workdir`` and return the aggregate verdict.

src/lra/contracts/verify.py:55

↓ 1 callersMethodviolations

Return ownership violations for a set of changed ``paths`` by ``writer``.

src/lra/coordination/ownership.py:88

↓ 1 callersMethodwrite_file

(self, relpath: str, content: str)

src/lra/execution/sandbox_e2b.py:41

Method__init__

(self, *, ceiling_usd: float, max_cycles: int)

src/lra/governor/governor.py:25

Method__init__

(self, *, threshold: int = 3)

src/lra/governor/governor.py:62

Method__init__

(self, *, dsn: str, embedder: Embedder)

src/lra/memory/semantic_pg.py:23

Method__init__

(self, embedder: Embedder)

src/lra/memory/skills.py:40

Method__init__

(self, embedder: Embedder)

src/lra/memory/semantic_memory.py:31

Method__init__

(self, k1: float = 1.5, b: float = 0.75)

src/lra/memory/hybrid.py:26

Method__init__

(self, *, semantic: SemanticIndex, bm25: BM25Index | None = None)

src/lra/memory/hybrid.py:78

Method__init__

(self, model_name: str = "BAAI/bge-reranker-v2-m3")

src/lra/memory/rerank.py:30

Method__init__

(self, dim: int = 256)

src/lra/memory/embeddings.py:23

Method__init__

( self, *, api_key: str, model_name: str = "voyage-3", dim: int = 1024

src/lra/memory/embeddings.py:45

Method__init__

(self)

src/lra/memory/episodic.py:13

Method__init__

(self, *, logger_name: str = "lra")

src/lra/obs/events.py:49

Method__init__

( self, *, api_key: str, model_name: str, default_max_tokens: int = 40

src/lra/model/claude.py:29

Method__init__

(self, providers: list[ModelProvider])

src/lra/model/failover.py:16

Method__init__

(self, model_name: str = "stub-1", script: list[TurnResult] | None = None)

src/lra/model/stub.py:22

Method__init__

( self, *, base_url: str, model_name: str, api_key: str | None = None,

src/lra/model/openai_compat.py:19

Method__init__

(self, workdir: str | Path)

src/lra/state/mission_anchor.py:39

Method__init__

(self)

src/lra/verify/flaky_quarantine.py:13

Method__init__

(self, *, default_timeout_s: int = 1200)

src/lra/verify/verifier.py:18

Method__init__

(self, *, coverage_threshold: float = 0.6, timeout_s: int = 900)

src/lra/verify/trust_bootstrap.py:35

Method__init__

(self, root: str = ".lra/objects")

src/lra/persistence/object_store.py:30

Method__init__

(self, dsn: str)

src/lra/persistence/repositories.py:18

Method__init__

(self, dsn: str)

src/lra/persistence/repositories.py:50

Method__init__

( self, store: ObjectStore | None = None, *, threshold_bytes: int = _DEFAULT_T

src/lra/durable/codec.py:25

Method__init__

(self)

src/lra/durable/saga.py:24

Method__init__

(self)

src/lra/durable/workflows.py:42

Method__init__

(self)

src/lra/safety/egress.py:28

← previous next →101–200 of 438, ranked by callers