v0.2.7

MartinLoop 0.2.7

This patch refreshes the public root package surface and tightens release hygiene for the CLI and SDK package.

Changed

• refreshed the published README, npm metadata, and repo-facing release docs so they describe the same MartinLoop product surface
• added stricter release checks that block internal process language and stray non-doc artifacts before a public release can ship
• removed leftover release artifacts from the public docs tree so external readers only see docs intended for them

Why it matters

This patch does not change MartinLoop runtime behavior.

It makes the package easier to evaluate and trust because the repo, npm page, and release workflow now stay aligned on one public-facing story.

v0.2.6

MartinLoop 0.2.6

This release hardened the root CLI and SDK package.

Changed

• stronger blocking for destructive verifier-command patterns
• broader prompt and verifier-output integrity checks
• tighter secret redaction and safe-path validation
• more consistent model-aware budget pricing and grounding-cache handling

Why it matters

Governed runs are safer at admission time and easier to trust when budgets, verifier input, and scope boundaries matter.

mcp-v0.2.0

@martinloop/mcp v0.2.0

0.2.0 turns the Martin Loop MCP server from a governed execution lane into a small local cockpit for reviewing governed agent runs.

0.1.4 introduced the safe operator foundation: check the environment, preflight a contract, run a governed coding task, and inspect saved results. 0.2.0 keeps that contract intact and adds read-only review surfaces so MCP hosts can show what happened after a run: recent runs, one-run dossiers, individual attempts, verifier results, and guided review prompts.

The important safety boundary is unchanged: martin_run is still the only tool that can execute work. The new 0.2.0 additions are read-only inspection and review helpers.

What Changed From 0.1.4

Area	0.1.4	0.2.0
Environment checks	martin_doctor	unchanged
Run preflight	martin_preflight	unchanged
Governed execution	martin_run	unchanged write boundary
Basic run inspection	martin_inspect, martin_status	unchanged
Review cockpit	not included	run lists, dossiers, attempts, verifier results
MCP discovery	tools	tools, resources, resource templates, prompts

Tools

Existing tools:

• martin_doctor
• martin_preflight
• martin_run
• martin_inspect
• martin_status

New read-only cockpit tools:

• martin_list_runs lists recent governed run summaries from the local run store.
• martin_get_run returns a run dossier by loopId or latest.
• martin_get_attempt returns one attempt record by loopId and attemptIndex.
• martin_get_verification_results extracts verifier completion events.
• martin_run_dossier assembles summary, task, budget, attempts, and verification evidence for review.

Resources

• martin://runs/summary
• martin://runs/latest

Resource Templates

• martin://runs/{loopId}
• martin://runs/{loopId}/attempts/{attemptIndex}
• martin://runs/{loopId}/verification

Prompts

• martin_review_run
• martin_triage_failures

Upgrade Notes

• Existing 0.1.4 tool callers do not need to change their current calls.
• Hosts can opt into the new cockpit surface by listing tools, resources, resource templates, and prompts through normal MCP discovery.
• The npm package remains @martinloop/mcp.
• The MCP server name remains io.github.Keesan12/martin-loop.

Verification

The published package was verified from npm after release. The published smoke test confirmed:

• package version @martinloop/mcp@0.2.0
• all 10 tools are discoverable
• both resources are discoverable
• all 3 resource templates are discoverable
• both prompts are discoverable
• local run inspection, status, run listing, dossier generation, and governed stub execution work from the installed npm artifact

v0.2.5

MartinLoop 0.2.5

This release improved persisted-run review and MCP setup around saved MartinLoop records.

Added

• triage views for saved MartinLoop runs so operators can review persisted records faster
• compact proof resources so MCP hosts can inspect saved run evidence with smaller responses

Changed

• inspection flows now skip unreadable saved entries and continue with warnings
• packaged host setup guidance is clearer for Codex, Claude, Gemini, and generic MCP hosts

Why it matters

Teams reviewing governed runs get faster triage, smaller proof payloads, and a more reliable inspection path when stored run data is imperfect.

mcp-v0.2.5

@martinloop/mcp 0.2.5

This release improved the standalone MCP package for reviewing real MartinLoop runs.

Added

• inspection tools for run lists, attempts, verifier results, and dossiers
• triage output for ranking persisted runs by urgency and missing evidence
• compact resources and prompts for proof review, debugging, and release checks

Changed

• unreadable records can now be skipped with warnings instead of breaking the whole inspection flow
• docs, manifests, and release checks were aligned with the shipped MCP surface

Compatibility

• martin_run remains the only execution entrypoint
• all newly added MCP surfaces are read-only except martin_run

v0.2.4

MartinLoop 0.2.4

This release expanded prompt and discovery support for MCP hosts.

Added

• guided prompts for kickoff, triage, resume, proof review, and release checks
• richer discovery metadata for surfaced MartinLoop capabilities

Why it matters

Hosts integrating MartinLoop through MCP can discover the available workflows more reliably and guide users with less custom glue code.

v0.2.3

MartinLoop 0.2.3

This release introduced compact run receipts and dossiers.

Added

• receipt-style dossier flows for reviewing what happened, why a run stopped, and what to inspect next

Changed

• clearer summaries and next-step recommendations for context-constrained agents and human reviewers

Why it matters

MartinLoop became easier to use as a review surface, not just an execution wrapper.

v0.2.2

MartinLoop 0.2.2

This release added triage for persisted MartinLoop runs.

Added

• triage support for ranking saved runs by urgency and evidence quality

Changed

• better handling for unreadable saved run entries so review can continue with warnings

Why it matters

Operators can focus on the runs that actually need attention instead of opening each saved record by hand.

v0.2.1

MartinLoop 0.2.1

This release improved local MCP setup for common hosts.

Added

• local MCP install and config generation for Codex, Claude, Gemini, and generic wrappers

Changed

• clearer host-specific setup guidance so the generated config is easier to understand and apply

Why it matters

Teams using MartinLoop through an MCP host can get to a working local setup faster and with less manual config editing.

v0.2.0

MartinLoop 0.2.0

This release established the first public CLI path for governed coding runs.

Added

• doctor for local readiness checks before a run starts
• demo plus a no-spend proof path for trying MartinLoop safely
• dossier --latest for receipt-style evidence review after a run

Why it matters

MartinLoop became easier to evaluate from a cold start: install it, run the demo, inspect the receipt, and decide whether governed coding fits your workflow.