Install
$ agentstack add skill-sergekostenchuk-ui-ux-agent-skill-system-project-wiki-manager ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ● Environment & secrets Used
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Project Wiki Manager
This skill is the project-memory coordinator for the UI/UX Agent Skill System. It decides what should be captured, what must stay out, which wiki backend skill to call, and how downstream agents should consume the resulting context.
Operating Modes
init: create the project memory plan and required wiki artifact map before a large project starts.capture: capture sanitized decisions, artifacts, constraints, and blockers after a major decision or delivery step.context-pack: prepare a compact context pack for the next agent or future session.sync: update existing wiki pages after task-plan, design, SEO, Figma, or launch changes.audit: check whether project memory is stale, secret-bearing, duplicated, or missing important decisions.handoff: produce a final sanitized handoff with exact artifact paths and next owners.
Required Workflow
- Classify project scale:
small: one localized task; wiki capture optional.medium: several files, one workflow, or one design direction; capture only major decisions.large: multi-session, multi-skill, SEO/UI/Figma/Stitch/infra, migration, or launch work; wiki gate required.
- Read [references/wiki-gate-contract.md](references/wiki-gate-contract.md).
- Read [references/sanitization-policy.md](references/sanitization-policy.md) before writing or ingesting any artifact.
- If creating a new project memory map, use [assets/project-memory-map.template.md](assets/project-memory-map.template.md).
- Choose backend:
wiki-capturefor current conversation decisions.obsidian-wiki-ingest/wiki-ingestfor existing documents and reports.wiki-context-packfor a compact context slice for the next agent.wiki-queryfor answering from stored project knowledge.
- Record exact source artifact paths, decision date, owner role, evidence status, and excluded sensitive data.
- Never store secrets, tokens, cookies,
.envvalues, recovery codes, private screenshots, or private URLs unless the user explicitly approves that exact storage.
Capture Triggers
Require a wiki gate when any trigger is true:
- The project uses
senior-ui-ux-orchestratorplus two or more specialist skills. - A three-variant design, council decision, migration, Figma handoff, Stitch output review, launch plan, or publication release is accepted.
- A progress-screen user gate records an accepted design choice, approval, blocker resolution, or launch confirmation that future agents must remember.
- The project spans multiple sessions or has a TASK-PLAN with open tasks.
- A future agent will need context that is not obvious from the current file tree.
- The user explicitly says not to lose context.
Skip or defer capture for tiny one-off edits, throwaway experiments, or artifacts containing unsanitized secrets.
Artifact Contract
Capture only durable facts:
- project goal and audience;
- accepted and rejected decisions;
- source documents and exact artifact paths;
- UX/SEO/Figma/Stitch/implementation constraints;
- progress-screen decisions, user gates, and accepted comments after sanitization;
- validation evidence and skipped checks;
- open blockers and next owners;
- approved context packs for downstream agents.
Do not capture:
- raw tokens, passwords, recovery codes, session cookies, private keys,
.envvalues; - private URLs or screenshots without explicit storage approval;
- unverified claims as facts;
- large duplicate source text that should stay in the original artifact.
Reporting Shape
Wiki mode:
Project scale:
Source artifacts:
Backend skill/tool:
Captured facts:
Excluded sensitive data:
Context pack path:
Updated wiki pages:
Checks run:
Next owner:
Validation
When editing this skill, run:
python3 $CODEX_HOME/skills/.system/skill-creator/scripts/quick_validate.py $CODEX_HOME/skills/project-wiki-manager
python3 $CODEX_HOME/skills/senior-skill-architect/scripts/lint_production_skill.py $CODEX_HOME/skills/project-wiki-manager
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: sergekostenchuk
- Source: sergekostenchuk/ui-ux-agent-skill-system
- License: Apache-2.0
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.