Fresh-Continue (GokayG/Fresh-Continue) is an MCP server listed on the M8ven Trust Index. It scores 49 out of 100, grade D. It declares 44 tools. No publisher has claimed this listing.

D
Caution
49/100

Fresh-Continue

Local-first Windows project & life progress system with an MCP brain for AI continuity.

Caution. Specific findings reduced this grade. They are listed on the page. Grades reflect the full trust pyramid: code, verification depth, and reputation. New projects cap at C until adoption is earned.

How we verified

⚡ Live Monitored: not connected

Verified is a snapshot. Live keeps it current, and builds your track record.

⚡ Connect GitHub → continuous verification on every pushwhy connect →

Who stands behind it

GokayG

Source: github_repo_search

Is this your MCP?

Claim it to get a verified publisher badge, a free copy of our full audit findings, and direct contact for any high-priority issues we find. Or connect your repo for our deepest verification, Live Monitored: read-only, revoke anytime. What we access →

Install from

The grade above is for the source repository. Registries can serve a different version, so we mark the ones we were not able to read.

// key findings
⚠️
Tool annotations don’t match behaviour
22 read-only tools perform write/delete/exec — project_list (line 25: spawn(process.execPath, [path.join(ROOT, 'core', 'index.mjs')], {); project_get (line 25: spawn(process.execPath, [path.join(ROOT, 'core', 'index.mjs')], {); project_resume (line 25: spawn(process.execPath, [path.join(ROOT, 'core', 'index.mjs')], {)
⚠️
Tool descriptions don’t match what handlers do
15 tools describe read intent but their handlers mutate — project_list (line 25: spawn(process.execPath, [path.join(ROOT, 'core', 'index.mjs')], {); project_get (line 25: spawn(process.execPath, [path.join(ROOT, 'core', 'index.mjs')], {); area_list (line 25: spawn(process.execPath, [path.join(ROOT, 'core', 'index.mjs')], {)
No credential exfiltration, no sensitive file access, no obfuscation
Static analysis found nothing flowing your secrets to unexpected places.
🔐
You'll be asked for 1 credential: FRESH_CONTINUE_RELAY_TOKEN
These are read from process.env at runtime. Make sure you trust where they’ll be sent.
// tools this server exposes44 tools

These names and descriptions are the publisher's own, read from the source code. We print them as written. Our assessment is the findings above, not this list.

project_list

List all tracked projects with progress, status metrics, current active nodes and ready-next nodes.

project_get

Read the complete current project graph, dependencies, decisions, checkpoints and node state before continuing work.

project_resume

Return a concise "where you left off" card: current step, next step, last meaningful activity, checkpoint and attention items. Use when returning to a project after a gap or when the user asks where they were.

area_list

Read broad areas such as work, personal, social and wellbeing. Areas are viewing/classification lenses, not object types.

calendar_read

Read dated project steps, deadlines, reminders and life/context plans in one chronological projection without duplicating the underlying objects.

calendar_collisions

Find overlapping scheduled project work and explicit life/context plans across Fresh Continue. Use before committing a new time when a cross-project, social or personal conflict is plausible. Deadlines and notification reminders are not treated as occupied time.

people_context

Read people explicitly linked to plans/tasks/context and their open work or remembered plans. Use when the user asks what is going on with a person, what they owe/expect, or before changing a person-related plan.

inactive_list

Read items that intentionally left the active flow but were not deleted. Use when recalling abandoned ideas, cancelled meetings/work or deciding whether to revive something.

item_lifecycle_set

Change lifecycle without deleting history. Use cancelled for explicit "not doing / cancelled / will not happen" intent, archived for intentionally shelved/reference-only items, and active to revive/restore. Cancelled is not Done.

project_create

Create a new tracked project. Use project_plan_set afterwards to build its dependency graph.

project_plan_set

Replace/reconcile the project roadmap graph. Existing completed nodes with matching IDs remain completed. Dependencies drive Ready/Locked state automatically.

plan_change_preview

Simulate changing a node schedule/deadline and show downstream dependencies, equivalent shifted dates, calendar collisions and pending plan reviews. Use before meaningful postponements or replanning; this tool never writes changes.

task_update

Update a roadmap node after real progress. Dependency locks are enforced by the core service.

project_context_add

Attach durable project context that is not necessarily code: notes, deliverables, references, URLs, metrics/KPIs, design decisions, user feedback, published assets, content ideas, research findings or other evidence. Use this for non-software projects as well as software projects.

project_context_list

Read project notes, deliverables, references, metrics and other evidence independent of Git or a local workspace.

project_add_decision

Record a meaningful architectural/product decision and its reason so later AI sessions do not rediscover or contradict it.

project_checkpoint

Create a checkpoint after a meaningful work session with summary, blockers, decisions and concrete next actions.

agenda_today

Read overdue, today, upcoming, recurring check-ins and reminders across all projects. Use for “bugün ne var?”, agenda, daily planning, deadline review and life-goal follow-up.

memory_radar

Find plan confirmations, stale open work, deadline risks, unscheduled active steps, waiting items, conversational flags, loose life commitments and people/entity context across Fresh Continue. Use to answer “what am I forgetting?”, during periodic reviews, or before replanning.

weekly_review

Prepare a calm weekly review from recent activity plus the memory radar: changed plans, stale work, unscheduled items, deadline risks and loose commitments. This is a review aid; it does not alter plans automatically.

project_reminder_add

Create a Windows-notification reminder for a project or roadmap node. Supports one-time, daily, weekly and monthly repetition.

project_checkin_add

Record a daily/weekly progress check-in for recurring goals, habits, agenda items or long-running roadmap nodes without permanently marking the node Done.

context_signal_record

Capture a concrete fact/change/commitment/cancellation/constraint from conversation when it can affect future plans. Link it to existing project steps and describe the impact. Conflicts/changes create a confirmation review instead of silently altering the plan.

context_signals_find

Look up recent active future-affecting facts, plans and commitments, optionally by person/entity, before deciding whether a new conversation detail conflicts with something already remembered.

context_reviews_pending

Read unresolved “is this plan still valid?” reviews created when later conversation may conflict with future steps.

context_review_resolve

Store the user’s decision about whether an earlier plan still applies. This closes the associated review reminder. Update the task separately only when the user chose to change it.

context_signal_invalidate

Use when a previously stored future-affecting fact is explicitly replaced by newer information. Keeps history instead of deleting it.

attention_flag_explain

Trace a conversational reminder back to its stored origin: project/node, context signal, plan review, creator/source and timestamps. Use when the user asks "why are you reminding me about this?" or doubts the memory.

attention_check

Cheap per-turn gate for ambient Fresh Continue behavior. Returns at most one eligible conversational reminder candidate plus the total due count. Call once on each user turn when Fresh Continue is available; do not mark anything surfaced merely by checking.

attention_flags_due

Read reminder flags that are currently eligible to be mentioned in conversation. This is intentionally topic-independent. Reading does NOT reset a flag. If you actually mention one to the user, call attention_flag_surfaced afterwards. If the user does not answer, leave it active.

attention_flag_add

Create a persistent conversational flag such as “ask me what happened with X later”. Flags survive unrelated conversations until explicitly resolved, and may be attached to a project/node or be global.

attention_flag_surfaced

Call ONLY after the reminder was actually stated to the user in conversation. This starts the cooldown but does not resolve the flag. If the user does not answer, the flag remains active and becomes eligible again later.

attention_flag_action

Apply the user response to a flag. Use resolve only when complete/not needed; snooze when the user says later; keep leaves it active. No user response means do not call this tool.

attention_flag_update

Change flag text, importance, cooldown, target project/node or earliest reminder time without closing it.

project_status_refresh

Use this when the user asks “proje ne alemde?”, “durumu güncelle”, “nerede kaldık?” or similar. It returns the current roadmap plus workspace/Git evidence and a reconciliation prompt. After reading it, inspect the actual workspace with your available tools, update stale nodes with task

project_summary

Return a concise human-readable summary of progress, active work, ready work and blockers.

project_handoff

Generate a comprehensive prompt that another AI/session can use to continue exactly where the project stopped.

project_activity

Read recent audit events to understand what changed and in which order.

relay_attention_check

Very lightweight per-turn check for due conversational reminders or pending plan confirmations when the local PC may be offline.

relay_status

Shows whether the local Fresh Continue PC has recently synced and how many offline events are waiting.

relay_context_record

Use only when a concrete statement can affect a future plan while the local Fresh Continue MCP is unavailable. The event will reconcile into the local app when the PC returns.

relay_flag_add

Create a conversational reminder in the always-on relay while the local PC is unavailable. It becomes usable immediately and syncs to the local Fresh Continue app later.

relay_flag_surfaced

Call only after you actually mention a relay reminder to the user. Starts its cooldown and queues the same state for local sync.

relay_flag_action

Resolve, snooze or keep an ambient relay reminder while the PC is offline; the action is later applied to local Fresh Continue too.

// environment variables
To run this server yourself, you supply these values. They go in your own MCP client configuration and stay on your machine. The secret label means the value is sensitive, not that the server mishandles it.
configFRESH_CONTINUE_CORE_URL
configFRESH_CONTINUE_HOST
configFRESH_CONTINUE_PORT
configFRESH_CONTINUE_RELAY_HOST
configFRESH_CONTINUE_RELAY_PORT
configFRESH_CONTINUE_RELAY_STATE
🔐 secretFRESH_CONTINUE_RELAY_TOKEN
configFRESH_CONTINUE_STATE_PATH
Deployment configuration, supplied by whoever hosts the server. Users are not asked for these.
deployPORT
// quality suggestions

All four hints declared on every tool

44/44 tools missing one or more hints — project_list (missing: openWorldHint); project_get (missing: openWorldHint); project_resume (missing: openWorldHint), +41 more. OpenAI's directory rejects tools where any of the four hints are missing or non-boolean.

For every tool, set all four hints (readOnlyHint, destructiveHint, idempotentHint, openWorldHint) to explicit true/false values that match the handler’s actual behaviour.

Tool annotations match behaviour

22 read-only tools perform write/delete/exec — project_list (line 25: spawn(process.execPath, [path.join(ROOT, 'core', 'index.mjs')], {); project_get (line 25: spawn(process.execPath, [path.join(ROOT, 'core', 'index.mjs')], {); project_resume (line 25: spawn(process.execPath, [path.join(ROOT, 'core', 'index.mjs')], {)

Either remove the readOnlyHint:true annotation, or remove the write/delete call from the tool handler.

Descriptions match behaviour

15 tools describe read intent but their handlers mutate — project_list (line 25: spawn(process.execPath, [path.join(ROOT, 'core', 'index.mjs')], {); project_get (line 25: spawn(process.execPath, [path.join(ROOT, 'core', 'index.mjs')], {); area_list (line 25: spawn(process.execPath, [path.join(ROOT, 'core', 'index.mjs')], {)

Rename the tool, rewrite the description, or move the side-effect into a separate clearly-named tool.

Tool inputs are validated

40/44 tool handlers declare input schemas (91%)

Declare an inputSchema with zod/joi/yup on every tool definition.

License file

No license file

Add a LICENSE file (MIT, Apache-2.0, etc.).

Tests exist

No test files found

Add tests that exercise each declared tool.

Tool description accuracy

14 tools have description/behavior mismatches: project_list: description implies read-only but handler writes/deletes/executes; project_get: description implies read-only but handler writes/deletes/executes; area_list: description implies read-only but handler writes/deletes/executes

Update tool descriptions to accurately reflect all capabilities — especially write, delete, or execute operations.

Claim the listing to review these findings one by one and send us a correction where you disagree, straight to the team. Claiming also means we tell you when the grade moves, and reach you first if we find anything urgent.

// full audit trail
The findings above are the summary. The full trail, every check we ran, each deduction, the network hosts observed and the dependency advisories, goes to verified publishers, along with an alert whenever a new one lands. Verified publishers can also review each finding and dispute it in one click. Publisher corrections have sharpened several of our checks this month, because the maintainer knows the codebase better than any scanner.
// improvement guidance — verified publishers only
We have 7 concrete improvements we can share with the publisher of this MCP. Each comes with specific guidance to raise the trust score.
// embed badge in your README
[![M8ven Verified](https://m8ven.ai/badge/mcp/gokayg/fresh-continue?variant=verified)](https://m8ven.ai/mcp/gokayg/fresh-continue)
Shows verification status without the grade. Want the grade badge instead? Remove ?variant=verified from the URL.
commit: d2e54fd7a2fe2b3df28a644db27cd82d568adc0f
code hash: 692e69b571fd6a672d33ad2865bcfd5e5497a3a560723ba704d0ab305175806c
view raw JSON →
Check MCPs from inside your assistant
Tool Check · MCP

Vetting this one by hand? Tool Check is an MCP that scores other MCPs. Add it once and ask Claude, ChatGPT, or any MCP client to grade a server, surface CVEs, check the publisher, and suggest safer alternatives — before you install.

https://m8ven.ai/api/mcp/tool-check
check_toolsearch_toolscompare_toolsrecommend_alternativescheck_publisherreport_concern
How to add it →Free · no account needed · works in any MCP client